Full text of "Road pricing : the economic and technical possibilities"

See other formats

MINISTRY OF TRANSPORT

ROAD PRICING:

The Economic and
Technical Possibilities

Report of a Panel set up by the
Ministry of Transport

LONDON

HER MAJESTY’S STATIONERY OFFICE
1964

Printed image digitised by the University of Southampton Library Digitisation Unit

FOREWORD

Economists have claimed that considerable net benefits could accrue to the
nation if vehicle owners had to pay higher charges or taxes when they used
congested roads than when they used uncongested ones, without there necessarily
being any change in the total motor taxation paid by them. These charges
would be in the nature of prices for using the roads, the prices varying from one
place and time to another according to the costs — notably the congestion costs —
involved in driving in a particular area at a particular time. The Ministry of
Transport set up a panel, under my chairmanship, to make a preliminary
examination of the technical feasibility of collecting such taxes, and to consider
some of the economic impUoations. The members of the panel have served in
an individual capacity and their report does not necessarily express the views
of their respective organisations.

I wish to emphasize that the panel have considered only how various forms
of taxes could be levied so as to differentiate between more and less congested
roads; it was not considered whether such a charge should be accompanied by
a change in the total revenue paid by vehicle owners. However, in order to
calculate some of the effects of such a system it was necessary to make some
assumptions about the distribution of the revenue and, for this pirrpose, the
panel assumed that the introduction of road prices would be accompanied by
a corresponding reduction in existing taxes so that the motoring population
as a whole would pay no more than it would otherwise have done.

After preliminary study of various proposals for implementing such a
change the panel have concluded that practical pricing methods could probably
be devised. Some of the information necessary for an accurate estimate of the
likely effects is not available, but it is tentatively estimated that the measurable
net gain to the community from the higher speeds consequent on the reduction
and reallocation of traffic would be about £100-£150 million per year under
present conditions. This figure takes no account of the non-measureable
benefits such as reduction of noise, fumes and frustration.

The full implications of such a revolutionary change are wider than the
panel’s field of enquiry. Inevitably some persons and firms would pay more
than they do now, while others would pay less. The panel have recommended
that such matters be investigated simultaneously with the development of
any pricing equipment that may be decided on.

In considering road pricing as a means of regulating traffic congestion the
panel have made the point that pricing by itself cannot produce a ‘cure’ for
congestion. The proposal to charge for use of congested roads should not be
regarded as an alternative to new and better roads; it is rather a means of
obtaining better value from the roads that already exist and from those that
are yet to be built.

R. J. Smeed
Chairman

B “

Printed image digitised by the University of Southampton Library Digitisation Unit

MEMBERS OF THE PANEL

Dr. R. J. Smeed, B.Sc., Ph.D., F.S.S. Road Research Laboratory

(Chairman)

W. F. Adams, M.I.C.E., A.M.Inst.T. Ministry of Transport

Dr. M. E. Beesley, Ph.D., B.Com.

R. H. Bird, B.A.

C. D. Foster, M.A.

W. O. Jennings, B.Sc.

G. J. Roth, M.A.,B.Sc.(Eng.)

Professor A. A. Walters, B.Sc.(Econ.),
J. G. Wardrop, B.A., F.S.S.

C. B. Winsten, M.A., F.S.S.

M. Thomson, B.Sc.(Econ.), F.S.S.
(Secretary)

London School of Economics

Ministry of Transport

Jesus College, Oxford

National Physical Laboratory

Department of Applied
Economics, University of
Cambridge

.S.S. University of Birmingham

Road Research Laboratory

Ministry of Transport and
Imperial College, London

Road Research Laboratory

Printed image digitised by the University of Southampton Library Digitisation Unit

TABLE OF CONTENTS

Section Page

Introduction

The Purpose of Road Pricing

2.2

Congested Areas

2.3

Uncongested Areas

Operational Requirements of a Road Pricing System

The Current Method of Charging

Other Indirect Methods of Charging

5.1

Differential Fuel Taxes

5.2

Parking Charges

5.3

Differential Licences

5.4

Other Indirect Congestion Taxes

The Feasibility of Direct Charging

6.2

Off-vehicle Recording Systems

6.3

Driver-operated Meters

6.4

Automatic Meters

6.5

Zone Marking

6.6

Supervision and Enforcement

6.7

Conclusions on the Feasibility of Direct Charging

The Prices

Economic Consequences of Direct Charging

8.1

The Overall Benefits

8.2

The Effect on Incomes

8.3

Private Motorists

8.4

Public Transport

8.5

Revenue

8.6

Effect on City Development

8.7

Investment Criteria

Recommendations for Further Work

Summary and Conclusions

Page

References

Appendices

Note on Marginal Cost Theory and Road Pricing

Calculation of Marginal Cost

Estimate of Economic Benefits from Direct Pricing

Benefits from Parking Taxes

Differential Fuel Taxes: Fuel Fetching Journeys

Printed image digitised by the University of Southampton Library Digitisation Unit

REPORT OF THE PANEL

1. INTRODUCTION

1.1 We were set up by the Ministry of Transport to study and report on the
technical feasibility of various methods for improving the pricing system
relating to the use of the roads, and on relevant economic considerations. We
were not required to consider the social and political aspects of any changes
which we might study; and, although it is difficult to extricate the technical and
economic issues altogether from these other matters, we have sought to avoid
the major social and political issues inherent in our subject. We are, however,
fully aware of their importance, and we would not wish our approach to the
subject to be misinterpreted on account of our limited terms of reference.

1.2 We began work in August 1962, since when we have held nine full meetings
and about 60 other meetings. We have considered over 30 papers on different
aspects of the subject, mostly prepared by our own members. We have
consulted 16 companies and research organisations for expert advice on
technical matters, and several firms have offered detailed proposals for new
types of pricing equipment which we have submitted in confidence to the
Ministry of Transport. We recognise that much further development may be
possible, some of it by firms which we did not approach in the course of our
studies. But we have at this stage given brief descriptions, in Section 6 of our
report, of those schemes which we have considered and think have technical
merit.

1.3 At the end of our report are five Appendices prepared for the Panel by
the Road Research Laboratory. We considered that they should be published
to provide a basis for further discussion.

2. THE PURPOSE OF ROAD PRICING

2.1 In 1962 motor revenues totalling £713 million were collected by means of
fuel tax, annual licences and purchase tax. These taxes are regarded by
vehicle owners as a charge for using the roads, although to what extent they
serve as a price system is debatable. Our task is to consider other methods of
charging whereby the price system may be usefully employed. One of our
main objects will be to see if the use of appropriate pricing methods can help
to lessen the problem of congestion. The existing taxes apply generally to the
whole road network, and in the final analysis it may well be necessary to
consider a pricing structure for the network as a whole, but we shall here
devote our attention mostly to the problem of congested areas and we shall
examine some pricing methods for specific use in such areas. We do not,
however, wish to give the impression that pricing can by itself produce a cure
for congestion. The need for investment in urban highways will still remain.

Printed image digitised by the University of Southampton Library Digitisation Unit

Congested areas

2.2.1 Congestion may be said to occur when one road user impedes the
movement of another. On highly congested roads one person may cause
inconvenience and delay to many others. For example, cars slow each other
down, pedestrians slow down cars, and fast-moving vehicles delay pedestrians
who wish to cross the road, or make them walk a long way to do so. These
obvious facts have led to many regulations and systems of control designed
to facilitate the movement of all kinds of traffic; for example, one-way
streets, parking restrictions, zebra crossings and cycle tracks.

2.2.2 Regulations of this kind often raise the question as to which of the
various conflicting demands on the use of the roads should be satisfied and
which should be restrained. For example, when a decision has to be taken
whether or not a road should be designated a clearway, how does one weigh
the interests of residents, who wish to park their cars and receive deliveries
of goods and so on, against the interests of vehicles passing through?
Before such regulations are imposed, assessments are necessary of the various
losses and benefits likely to result. Changes in the pricing methods by
which people are charged for using the roads also cause a mixture of losses
and benefits, the net result of which has to be assessed before one can say
how valuable the changes are. When dealing with prices, however, the
problem of assessment is eased by the fact that some road users can themselves
give some information about the value they attach to their respective uses by
their willingness to pay the price. Different people attach very different
values to their various uses of the road; and the value they attach to a
particular use may vary greatly from one time to another.

2.2.3 Prices charged for using the road generally have the effect of restraining
the amount of use that people make of the roads. Two questions that
immediately arise are: how far should use be restrained, and should some
traffic be restrained more than other traffic ? A useful guiding principle is
that journeys should not be made if they are valued at less than the costs or
losses they cause to other people; similarly, journeys should not be restrained
if they are valued at more than the costs they cause. If this underlying
principle is neglected, a waste of resources is likely to result.

2.2.4 How, then, can one compare the value of a journey with its cost?
The money value attached by a road user to his journey is the price which he
would be willing to pay for it; and the money value of the costs caused to
other people is what they would pay to avoid these costs. The extent to
which we are prepared to accept that these two money values can be compared
depends largely upon (a) how far we consider that the distribution of wealth
and income between the parties involved is acceptable,* and (b) how nearly
the prices of alternative means of transport are similarly based on cost.
Having regard to these things, when a pricing system is used on the roads,
a useful general rule upon which to base prices is that the road user should
pay a sum equal to the costs he imposes upon others. Subject to the above

* The parties include firms and other organisations as well as individuals.

Printed image digitised by the University of Southampton Library Digitisation Unit

qualifications and given the manner in which the revenue from the pricing
system is to be disbursed (see Section 8), the most beneficial way of obtaining
it is likely to follow upon the application of this rule.*

2.2.5 The introduction of particular pricing systems must therefore depend
on a careful calculation of the costs and benefits arising from them, both in
the short and in the long run. These would need to be compared with the
costs and benefits of alternative schemes, such as regulatory schemes which
might forbid the use of some kinds of vehicles in certain city centres.

2.2.6 The costs that a vehicle user imposes on other people when he makes
a journey may be divided into three classes:

(a) Road costs: the wear and tear on the road, and the cost of lighting,
traffic control, etc. According to available estimates, for the great
majority of vehicles these costs are unUkely to exceed a Jd. for each
mile travelled in towns.(

(b) Congestion costs: the costs of delay to the rest of the traffic. For
convenience we divide these into :

(i) the costs to cyclists, pedestrians wishing to cross the road, and
other road users apart from motor vehicle users. We have as
yet no estimates of these costs.

(ii) the costs to other motor vehicle traffic. The cost of delay includes
higher labour costs, loss of people’s time, higher fuel and running
costs and lower utilisation of vehicles and their loads. Research
on some of the important components of these costs has been
carried out at the Road Research Laboratory. Estimates for
central London, for example, show that the congestion costs
imposed by a typical car on other vehicles rise from 4d. a mile
at traffic speeds of 20 mile/h to 2/- a mile at 12 mile/h and 6/-
a mil e at 8 mile/h. t The costs imposed by heavy vehicles are
often two or three times higher.

(c) Other social losses: the losses imposed on the general community —
pedestrians, motorists, residents and aH others affected by the accident
risk, noise, fumes, dirt and other undesirable effects. We have no
quantitative information about the magnitude of these costs.

2.2.7 Clearly, the costs caused by road travel in towns can be high. The
possibilities of restraining people from adding to these costs without sufficient
reason may be illustrated by a simple analogy. Suppose a man works a
small ferry boat continuously across a river and that it costs 4d. a minute
to run the boat, including the man’s wages. Suppose that the boat can
seat 10 passengers comfortably but can actually hold as many as 20 passengers.
Suppose that each succeeding passenger increases the load so as to add half
a minute to the time of the crossing. If there were six passengers in the boat

* See Appendix 1 for a fuller discussion of pricing principles. See also paras. 2.2.7 to 2.2.12
below.

t These figures are tentative and more research is needed, especially on the valuation of
leisure time. See Appendix 2.

Printed image digitised by the University of Southampton Libr^ Digitisation Unit

and a seventh joined them he would increase the ferryman’s costs by 2d.
and he would also cost each of the other six passengers half a minute’s delay
which might be worth, say, |d. each or 3d. in all. If there were 17 passengers
each person would in effect be adding 2d. to the ferryman’s costs and would
also, by delaying 16 other people, be adding 8d. to their personal costs and
appreciable discomfort on their journey. The ferryman might cover his
costs if he let each passenger on for only 4d. but at this price the time and
comfort of the majority of passengers might be sacrificed for some journeys
which were not worth so high a sacrifice. There are several possibilities of
improving this situation ;

(1) Various regulations might be introduced. For instance, it might be
decided that not more than 15 passengers should be allowed on the boat,
on the grounds that more than this number created too much delay and
discomfort. The problem of which passengers should be allowed on could
be dealt with on the principle of “ first come, first served ”, or alternatively
some system of priority might be devised, giving preference perhaps to those
living nearest the ferry and depending on it most.

(2) The passengers could perhaps bargain amongst themselves. They
might all be prepared to pay something to persuade one or more of the
others to leave the boat. Some of them, less eager than the others to make
that particular trip, might agree to accept a payment and take a later boat,
say, or walk round by the nearest bridge. Thus all the passengers — but not
the ferryman — would be better off.

(3) The ferryman could raise his price. A higher price could remove
those passengers not prepared to pay the cost of time, money and comfort
that their journeys would cause to the ferryman and the other passengers.
As the numbers were thus reduced the associated costs would also be reduced,
and if the price to the remaining passengers was set so that it just covered these
costs this would ensure that journeys were not made if they were valued at
less than the costs they caused. In this way, there would be an aggregate
gain. But how this gain was distributed between the passengers, the ferryman
and ultimately the rest of the community would depend upon many factors
and could be affected — especially in the long term — by all sorts of con-
sequential changes.

2.2.8 In each of these three methods there is likely to be a different result
in the distribution of income and advantage between the passengers, the
ferryman and the wider community. The use of the roads involves similar
problems to those of the ferry. A wide variety of regulations are possible.
Alternatively, the bargaining analogy of 2.2.7 (2) would be if potential users
of roads in a particular town paid each other to change their modes of
transport. For example, if at some time of day buses were found to be a less
congestion-forming method of transport than cars, taxes on cars might be
used to improve the service for bus passengers, i.e. the oar-owners who were
almost indifferent between cars and buses would be paid by other car-owners
to transfer to buses.

2.2.9 But there is no obvious system of direct bargaining by which people
can persuade each other — ^pay each other — to stay off the road. A price
system has to be imposed by some authority. The price mechanism enables

Printed image digitised by the University of Southampton Library Digitisation Unit

people to assess their needs and to base their choice on the price that
confronts them. For example, if one compares the pricing of cars with the
banning of their use in city centres, a price system would enable people who
strongly wanted to use the roads to do so if they were prepared to pay the
price, whereas a ban might keep out some people who could obtain great
benefit from their journeys and would readily pay a good deal for them.

2.2.10 In order to fix the prices the authority should predict the effects of
different prices by measuring the costs and benefits to road users and others
affected by congestion. Very important amongst these losses from con-
gestion are losses of time, and therefore the pricing authority must measure
time spent, time saved, etc. for the different components of traffic, and also
establish what valuations are put upon time and the other costs by the
different users.

2.2.11 Some calculations of the economic benefit of introducing a particular
system of charges for the use of roads in towns have been made at the Road
Research Laboratory. The method used and the assumptions made are
described in Appendix 3. Costs have been estimated, and a few different
assumptions have been made about the restraining effect of different prices.
Appropriate price levels have then been estimated, from which the overall
benefits have been assessed at between £100 and £150 million a year; they
refer only to users of motor vehicles and exclude all benefits to pedestrians,
cyclists and residents, and also all benefits resulting from the reduction of
fumes, noise, dirt, vibration and so on, and any benefits which might arise
outside towns. The figures are necessarily tentative and exclude social
considerations, but they are believed to indicate the orders of magnitude.

2.2.12 Any such changes in the system of road charging would have different
effects in different parts of the country and on different sections of the
community, i.e. they would redistribute income. The effect on different
sections of the community would depend crucially upon the way in which
the proceeds of high road charges in towns were used. There are many
possibilities. For instance, the proceeds could be divided in various ways
between the local and central financial authorities; or they could be used for
improving the road system, or for subsidising public transport in certain
towns; or simply as general revenue. Whatever policy was adopted, the
economic implications and the measurement of redistributive effects would
provide a most important subject for future research. This, however, is a
question which lies largely outside our present terms of reference. It would
be wrong for us to express any view as to whether the current level of motor
taxation is too high or too low and we do not wish to imply any such view.
But we do wish to show that, for any given level of motor taxation, the
economic consequences of one method of collection may be preferable to
those of another. In order to demonstrate this, we have made calculations
based on the current level of motor taxation, thus assuming that the additional
proceeds from the new method of road charging would be returned in full to
the motoring community by a reduction of existing taxes. But this assump-
tion describes only one possible situation, not necessarily the best, and
should not be regarded as an argument against either lower taxes or higher
taxes.

Prints image digitised by the University of Southampton Libi^ry Digitisation Unit

2.2.13 The final choice amongst all schemes, in whatever proportion they
depend on price or regulation, must be based on their expected consequences.
The final assessment must take account of both economic and other factors.

Uncongested areas

2.3.1 Our remarks on the principles underlying road pricing in congested
areas apply also in uncongested areas, with the important difference that
there is, by definition, no cost of congestion to be charged for. This means
that the cost of travelling on uncongested roads is not only much lower but
also much less variable. That is to say, the total cost — to the vehicle-owner
and all other parties — caused by travelling on uncongested roads normally
varies little from one mile to another, except where important social costs
arise such as damage or disturbance to beauty spots or village communities ;
whereas on congested roads the cost of travel varies greatly according to the
degree of congestion experienced and the type of environment encountered,
e.g. shopping centres, residential quarters, factory areas and so on.

2.3.2 Because of the relative lack of variability in the cost of using un-
congested roads we think that the advantages to be gained there by changing
the existing method of taxation are much less than in congested areas. The
level of the fuel tax may well affect the amount of use made of roads in
uncongested areas ; we have not studied this question. Nor have we studied
the costs to the community resulting from the use of uncongested roads.
We feel that estimates of these costs should be obtained, so that consideration
could be given to possible changes in the taxation paid in these areas.

2.3.3 Our concern, however, is not with the level of taxation but with the
method of charging, and since there is no obvious and urgent need for a
change of method in uncongested areas we shall not consider specifically the
pricing of these areas. Most of the pricing methods we discuss will, in fact,
be inapplicable to rural areas.

2.4 In the following pages we have first set out what we regarded as the
operational requirements of a road pricing system. In the light of these
requirements we have then attempted to assess the technical and economic
merits of various pricing methods, commencing with the fuel tax and annual
vehicle licence, and continuing with the differential fuel tax, differential licences,
parking taxes, poll tax on employees, vehicle recording systems, driver-operated
meters and automatic meters. When reading through these sections it will be
found helpful to refer to the diagram on page 34.

Printed image digitised by the University of Southampton Library Digitisation Unit

3. OPERATIONAL REQUIREMENTS OF A
ROAD PRICING SYSTEM

3.1 Having considered the relevant principles underlying the pricing of the
roads we have compiled a list of operational requirements that are necessary
if these principles are to be closely followed. This list forms the basic specifica-
tion for a road pricing system and provides the starting point for our technical
investigations. It consists of nine requirements which we regard as important
if the method is to be workable and is to come tolerably close to its objectives ;
and a further eight requirements which are also considered desirable.

3.2 The most important requirements of the system are as follows: —

(1) Charges should be closely related to the amount of use made of the
roads.*

(2) It should be possible to vary prices to some extent for different roads
(or areas), at different times of day, week or year, and for different classes of
vehicle.

(3) Prices should be stable and readily ascertainable by road users before
they embark upon a journey.

(4) Payment in advance should be possible, although credit facilities may
also be permissible under certain conditions.

(5) The incidence of the system upon individual road users should be
accepted as fair.

(6) The method should be simple for road users to understand.

(7) Any equipment used should possess a high degree of reliability.

(8) It should be reasonably free from the possibility of fraud and evasion,
both deliberate and unintentional.

(9) It should be capable of being applied, if necessary, to the whole country
and to a vehicle population expected to rise to over 30 million.

3.3 The requirements which we consider also desirable, but not so important,
are as follows: —

(10) Payment should be possible in small amounts and at fairly frequent
intervals, say amounts not exceeding £5 and intervals not exceeding one month.
This does not exclude payment in larger amounts where preferred.

(11) Drivers in high-cost areas should be made aware of the rate they are
incurring.

(12) At the same time the attention of drivers should not be unduly diverted
from their other responsibilities.

(13) The method should be applicable without difficulty to road users
entering from abroad.

(14) Enforcement measures should impose as little extra work on the
police forces as possible and should therefore lie within the capacity of traffic
wardens.

* The most obvious ways of measuring road use are in terms of time or distance, each of
which has its advantages and disadvantages. A study of this question is recommended
in para. 9.1.

Printed image digitised by the University of Southampton Library Digitisation Unit

(15) It would be preferable if the method could also be used to charge for
street parking.

(16) The method should, if possible, indicate the strength of demand for
roadspace in different places so as to give guidance to the planning of new
road improvements.

(17) The method should be amenable to gradual introduction commencing
with an experimental phase.

3.4 We have had to consider what is an acceptable cost for such a system.
We suggested above that the net benefits to the community from efficient road
pricing would be worth £100-£150 million a year if introduced as at present
(see Appendix 3); and these potential benefits could increase at a rate much
faster than the vehicle population itself. By usual investment standards a
very high cost, if necessary, would be justified by a return of this magnitude;
but, for practical reasons, we have set an arbitrary target of about £1 per
vehicle per year, including servicing of capital, and we have aimed at a capital
cost of under £5 per vehicle. Thus we have regarded as expensive those
schemes which, for a present population of 10 million vehicles, require an
initial capital expenditure of over £50 million or an annual administration cost
of more than £5-£10 million.

4. THE CURRENT METHOD OF CHARGING

4.1 Motor vehicles used on the roads are at present taxed by means of fuel
tax, annual licences and purchase tax. The money paid in these various ways
provides a contribution towards the general revenue of the state. It is not for
us to comment upon the desirability of drawing general revenue from this
particular source, nor upon the volume of such revenue, but it is of relevance
to our problem to note that these taxes, from the point of view of the vehicle
owner, stand as a crude scheme of prices for using the roads. We have,
therefore, gone briefly into the origins of the present tax system and have
considered how far it provides a satisfactory substitute for a price system.

4.2 After the decline of the turnpikes the cost of building and maintaining
roads was borne by local and central government. The taxpayer paid for the
roads, and the amount he paid was determined by the amount of road capacity
which the community decided to provide. The use of the roads was “ free ”
in that there was no connection between the amount of road capacity he paid
for and the amount he personally used.

4.3 With the advent of the motor vehicle large and costly road works were
required, and in the Budget of 1909 motorists were made to shoulder these
costs by a two-part system consisting of a fixed tax and a variable tax. The
variable tax was the petrol tax which served as a rough and ready way of relating
the charge to the amount of use made of the roads. But once a road was
built and in use, the cost of letting additional motorists use it was very low, so
that a petrol tax which covered only this cost would not have covered the
capital costs and other overheads. On the other hand, if the tax had been
raised to cover these fixed costs, motorists would have been needlessly
deterred from using roads which once built could accommodate additional

Printed image digitised by the University of Southampton Lit^ry Digitisation Unit

traffic at very little extra cost. The anmaal licence fee provided a second
source of revenue which, being unrelated to mileage, did not deter motorists
from using the roads except inasmuch as it may have deterred some people
from becoming motorists at all.

4.4 In its essentials the 1909 system of road taxation has remained unaltered
to the present day. The amounts have risen, of course, and parking charges
have been added in some places. Purchase tax on vehicles has also been
added. During the first world war and after 1926 contributions to general
taxation were also made and the current practice is to credit the whole revenue
from motor taxation to the Exchequer.

4.5 The annual administrative cost has been estimated as £3.6 million for
the collection of motor licence tax (including the issue of driving licences) and
under £1 million for the collection of fuel tax. The total cost of parking
controls is not known, but it is estimated that in parts of London (Croydon,
Finsbury, Holborn, St. Marylebone, St. Pancras and Westminster) the average
cost per meter (including wages, maintenance, administration and loan charges)
is about £41 per year.

4.6 We have called the existing system a two-part system. It is also an
indirect system in that both parts are related only indirectly to the use of the
roads. The fixed part stands at £15 per annum for a private car, with a steeply
rising scale for commercial vehicles. The variable part stands at 2s. 9d. per
gallon of fuel which represents a road charge of about Id. per mile for a typical
car, IJd. per mile for a medium diesel-driven lorry and nearly 3d. per mile for
a medium petrol-driven lorry.(2) These rates work out higher in congested
conditions. If the use of the roads is measured by the length of time spent
on the road, the rate paid by a typical car varies from about Is. 4d. per hour
in a congested area, assuming an average speed of 10 mile/b, to about 3s. 6d.
per hour on a good uncongested rural road. Parked vehicles pay nothing,
or some 6d. per hour if at a parking meter, but they are increasingly prevented
from occupying road space that is of much value to moving vehicles.

4.7 The 1909 road taxes resulted from important technical innovations which
created a new factor on the roads — ^the motor vehicle. Since then continuing
technical change combined with rapid social change have created another new
factor — ^widespread severe congestion. The method of charging fails to meet
the needs of the situation in several important respects. It fails completely
to discriminate between those situations in which congestion costs of road use
are high and those in which they are low, and we have already mentioned how
very large this difference can be. A charge of Id. per mile is higher than the
cost of using uncongested roads and far lower than the cost of using congested
roads and thus discourages the use of lightly-trafficked toads and encourages
the use of crowded roads, though it is not known by how much.

4.8 Secondly, the use of the fuel tax is a rather haphazard method of relating
the charge to the amount of use made of the roads. For example, a car with
a high fuel consumption rate pays more than a car of similar size with a low
fuel consumption rate, and a petrol-driven lorry more than a diesel-driven
lorry.

Printed image digitised by the University of Southampton Library Digitisation Unit

4.9 Thirdly, the universal use of a fixed annual charge appears ‘ a priori ’ to
have even less effect than the fuel tax in restraining congestion, although of
course it may serve other useful purposes.* Its continued use in some outlying
areas might, however, still be advisable as a means of raising revenue or con-
tributing towards the overhead costs of the road system.

4.10 We conclude that, given the immense growth in the number of vehicles,
the present taxation methods do not effectively restrain the use of the roads in
the right places at the right times and new methods may have much to contribute
in limiting the losses due to traffic congestion.

5. OTHER INDIRECT METHODS OF CHARGING
Differential Fuel Taxes

5.1.1 The current method of charging suffers from the inflexibility of
charging a fixed standard rate for fuel tax. We shall therefore consider
whether any benefit could be obtained by introducing differential fuel taxes.

5.1.2 It would be possible to levy fuel tax at different rates in different
areas, relating it to the amount of congestion in each area. The technical
feasibility of such a tax is not in doubt but its economic advantage would,
we think, be small. The effectiveness of the differential would depend on
the opportunities and incentives for avoidance, and on its ability to discrim-
inate in detail against journeys on congested roads.

5.1.3 Avoidance could occur in two ways : —

(a) taking on fuel while travelling on ordinary journeys through low-tax
areas;

(b) making special fuel-fetching journeys to low-tax areas.

Avoidance of type (a) could not be stopped and must be accepted as an
inherent weakness of the method. Avoidance of type (b) could be minimised
by the careful planning of area boundaries and differentials.

5.1.4 Let us assume that motorists w'ould merely fill their tanks and would
not take reserve cans or drums to fill in low-tax areas. They would have to
compare the gain from buying a tankful at a cheaper rate with the cost of
driving to the cheaper area and back again. This cost would consist of
vehicle running costs and the driver’s time. The arithmetic of the tax
is examined in Appendix 5, with the conclusion that fuel-fetching will be
unprofitable for most motorists if the differential is under Id. per gallon
per mile, and will be profitable for most motorists if the differential exceeds
2d. per mile. If a small amount of fuel-fetching is to be tolerated, the
differential could be as high as l|d. per gallon per mile. Thus, the maximum
workable differential between places 20 miles apart is roughly between
Is. 8d. and 2s. 6d. per gallon, and for places 30 miles apart it is between
2s. 6d. and 3s. 9d. per gallon, assuming that the petrol used on the fuel-
fetching journey is tax-free.

* Additional revenue may be required either to cover the cost of the road system (see
Appendix 1 , para. 6) or for general tax purposes.

Printed image digitised by the University of Southampton Libf^a^ Digitisation Unit

5.1.5 Quite large differentials are therefore possible, but only at intervals
which are bound to include much low-cost road in high-tax areas and much
high-cost road in low-tax areas.

5.1.6 There are two ways of setting areas: a ‘contour’ method whereby
the areas consist of narrowly spaced contours around congested centres;
and a zone method where the zones are large areas including whole con-
urbations. Zoning necessarily has the disadvantage of the boundary line
with a big tax differential between one side of the hne and the other. A
certain amount of fuel-fetching is therefore certain to occur in the fringe
area if the differential is to be at all effective; and the higher the differential
the larger will the fringe area be : witness the fuel-fetching by French motorists
from across the Swiss border where petrol is approximately 3 shillings per
gallon cheaper. With the contour method there should be no fuel-fetching
of any significance, if the differentials are kept within the workable limits
suggested above. But, as soon as they exceed these limits, fuel-fetching
begins to be profitable over large distances. With both methods, if the
workable limits are exceeded, fuel-fetching is likely to become widespread.

5.1.7 Differential fuel taxes are therefore subject to a fundamental limitation
on the size of differentials that are practicable between nearby areas. In a
large, less built-up country this limitation would be of far less consequence.
But in England it detracts seriously from the usefulness of the method.

5.1.8 Nevertheless a large measure of differentiation in the fuel tax is
possible. What economic benefits could be expected? The scheme com-
mends itself most obviously to conurbations, where a high tax could be
levied throughout most of the built-up area and could be reduced gradually,
as one moved away from the area, starting in the outer suburbs. If the total
yield from the fuel tax was kept at its present level, many smaller towns and
other congested areas would inevitably be included in cheap areas. Thus,
while congestion might be eased in the conurbations, it might be worsened
elsewhere.

5.1.9 In the conurbations, provided that avoidance of type (a) was not
very great — ^though it probably would be — and that illegal evasion, say by
black marketeering, was not too great, a fuel tax of Ss. per gallon would
undoubtedly help to reduce congestion. But this does not necessarily mean
that the tax would be beneficial. It would also discourage traffic in un-
congested areas and at uncongested times, and the losses arising from this
source might offset the gains in congested areas.

5.1.10 There are other difficulties, such as the effect on the fortunes of
individual garages and the waste of resources caused by fuel-fetching.
We cannot pretend that the assumption which we made for our calculations,
that motorists would not fill cans or reserve tanks, is very reahstic. Also
commuters as a class would find avoidance easy unless the high tax areas
were so large that they severely penalised a great deal of traffic in uncongested
areas. We conclude that differential fuel taxes by themselves are no solution
for the pricing problem. At most, they might assist other measures.

Printed image digitised by the University of Southampton Library Digitisation Unit

Parking charges

5.2.1 Parking restrictions and regulations can also be used to reduce
traiEc congestion on the roads, since they can be used to increase the
difficulty, or put up the price, of making a journey and thus reduce the
volume of moving traffic. We do not here give a complete appraisal of
these methods, but only discuss the use of parking charges. The aim of
such charges would he to reduce the number of parking places already
available, or to reduce the number actually in use at any one time.

5.2.2 Public authorities often control a large proportion of the parking
places at present available in congested areas, since these are usually on the
street. Using parking charges to control congestion implies that these
parking places would all be charged (as many of them are now) through
meters and that the charges would be raised to a level which would lessen
the number of journeys made in the congested area. If there are many
parking places, the price may well be so high that a high proportion will be
vacant much of the time.

5.2.3 If parking charges were high in pubhc parking places, then there
would be an incentive for the provision of commercial parking places, and
also for an increase in garages and parking places attached to offices,
factories and private homes. We have no estimate of how great this tendency
is likely to be, but if it is large enough to counter substantially the reduction
in public parking, then it in turn would have to be restrained. One possible
way of doing this would be to impose a tax on parking places, both com-
mercial and private. Such a tax, however, seems irrational and difficult to
justify to those users who could easily provide themselves with private parking
space, and this would make it difficult to enforce.

5.2.4 Those users who, despite the tax, did provide themselves with parking
spaces at their offices, homes, etc. would treat the charge as a fixed one,
independent of the number of journeys. For such users it would not
influence the number of journeys.

5.2.5 At metered parking places a charge is made whenever a vehicle
parks. Thus, in effect, a charge would be made each time a journey finished
at such a place within a congested area. With commercially provided
parking spaces, if these were taxed, the tax may well be passed on in a charge
to the customer, so that in either case there would be an extra charge imposed
for each journey.* Such charges would be independent of the length of the
journey, although they would, of course, often vary with the length of time
the vehicle is parked.

5.2.6 All the calculations to which we refer below assume a uniform charge
per journey, which for convenience we will call a ‘ parking tax ’.

5.2.7 Economic advantage. There is no doubt that in some places a
parkmg tax, rigorously enforced and set high enough, could deter large
numbers of vehicles from using the roads. It must be remembered however
that the economic objective of the policy is not simply to reduce the volume
of traffic, but to obtain the maximum benefit from the available road system.

* One could ensure that it was passed on by levying it
commercial providers of parking space.

as a sales tax on the revenue of

Printed image digitised by the University of Southampton Library Digitisation Unit

This is not achieved simply by forcing a certain amount of traffic from the
road; the aim must be to reduce the traffic which is not worthwhile in the
sense that it is not prepared to pay the real cost of its being there.

5.2.8 A parking tax is a crude way of charging vehicles for the use of the
roads, because it bears no relation to the distance travelled nor to the type
of roads used. Nevertheless, by placing a financial penalty on all vehicles
parking in the area, it must deter some of the least worthwhile traffic. The
questions then are : how far can a parking tax lead to the right volume and
composition of traffic? How much net benefit can it achieve?

5.2.9 Calculations have been made of the benefits which might be produced
by a parking tax in different sets of circumstances. These are described in
Appendix 4. Broadly they attempt to take account of the three main
deficiencies of such a policy. The first is that traffic which does not park
in the area, or which manages to avoid taxed parking places, is not merely
unaffected but is actually encouraged by whatever reduction the scheme
extracts from other traffic. Thus traffic wishing to use the area is kept out
to make way for traffic passing through the area. This difficulty may be
alleviated by specific road improvements and traffic engineering techniques.
Secondly, a vehicle coming in from outside the town, causing congestion all
the way, pays the same as a vehicle coming only a short distance and, because
the tax forms a smaller proportion of its total journey cost, is less likely to
be deterred by the tax. Thus the traffic that causes least congestion is kept
out to make way for the traffic that causes most congestion. Thirdly, a
tax that is efficient in the peak hours is likely to be too severe in the off-peak
hours. If the same tax has to be charged throughout the day a compromise
must be sought between an over-congested peak period and an under-
utilised off-peak period. A fourth factor, omitted from the calculations, is
the effect of the tax on vehicles parked in the area but not used during
congested hours.

5.2.10 The net economic benefits would be the gross benefits to remaining
traffic due to reduced congestion, less the losses to the traffic forced off by
the tax. It has been estimated that in Central London the best tax — equal
to about 6s. a day — would reduce traffic by about 11 per cent in the off-peak
period and by 7 per cent in the peak. As a result, off-peak speeds are
calculated to rise from 10.7 mile/h to 12.6 mile/h, and peak speeds from
9.8 mile/h to 11.1 mile/h. The real cost of travel (i.e. the cost excluding tax)
would go down by 12 per cent but the real savings (after taking losses into
account) would be about 7^ per cent or 1.5 pence per original vehicle-mile
(see Appendix 4).

5.2.11 This saving is comparatively modest; both the effect on speed and
the net benefits obtainable from the tax are about 40 per cent of what is
obtainable from direct pricing systems.

5.2.12 These estimates are based on present traffic demands and on a
30 per cent proportion of non-parking traffic. However, the demand for
uncongested road space, and hence the amount of non-parking traffic (which
is not deterred by the tax), may be expected to increase from year to year,
and the benefits from the parking tax would decrease. Thus with non-
parking traffic at 25 per cent, a tax of 6s. a day would yield a net benefit of

D 13

Printed image digitised by the University of Southampton Library Digitisation Unit

1.6d. per vehicle-mile; with non-parking traffic at 50 per cent the best tax
would be 7s. a day and would yield a net benefit of 1.3d. per vehicle-mile.

5.2.13 We conclude that a parking tax may, under certain conditions, give
some beneficial relief to traffic congestion, although it can never of itself
be regarded as an efficient solution. If the proportion of non-parking
traffic is large or susceptible to a big increase as soon as the state of con-
gestion permits, the tax will give little net benefit. This may have special
relevance for small areas of congestion where the proportion of through-
traffic is often large. In larger areas, on the other han^ if the traffic tends
to consist mainly of long and short-distance traffic, with little middle-
distance traffic, again the parking tax will give little net benefit.* If there
is a very pronounced peak traffic, this too will detract from the value of the
tax, except insofar as the proportion of parking traffic (e.g. commuters)
may be higher during the peak than during the off-peak.

Differential Licences

5.3.1 We discuss two forms of differential licence: first, the differential
annual licence suggested in 1952 by lames M. Buchanan and (independently)
in 1954 by Walters(3); and second, the differential daily licence. Under
both schemes it is proposed that the licences should be varied in two ways : —

(1) by area, differently priced licences being required for different areas,
and

(2) by time, different licences being required for different times of the
day.

5.3.2 Area -variations would require congested areas to be divided into
zones of different classes, for example, ‘ red ’ zones, ‘ blue ’ zones and
‘ yellow ’ zones. Areas of heavy congestion could be designated red, areas
of moderate congestion blue, and less congested areas yellow. Special
licences would be required in these areas. A yellow licence would give
access to yellow zones, a more expensive blue licence to blue and yellow
zones, and a stiU more expensive red licence to all three classes of zone.

5.3.3 Time variation would distinguish between day and night usage and
possibly between peak and off-peak usage. A zone might be declared
‘ red ’ at the peak, ‘ blue ’ during day-time off-peak hours and ‘ yellow ’ at
night; or more simply, ‘blue’ at peak hours and ‘yellow’ at all other
times.

5.3.4 Feasibility of the annual licence. To be effective, licences for heavily
congested areas would have to be (a) expensive— £100 a year might be the
level required, and (b) transferable, so that people could obtain a licence
for a short period and people holding a licence would have an incentive to
give it up when they did not have any strong reason for keeping it.

5.3.5 The price of an annual licence would be fixed at a level calculated to
limit the volume of traffic at peak times to a predetermined amount. The
success of the scheme depends upon the creation of a free market in licences.
The dealers in this market wonld probably include most garages, who would

* This point is brought out in Appendix 4.

Printed image digitised by the University of Southampton Library Digitisation Unit

buy licences from the authorities and hire them at market prices to the
public. Motorists would thus — according to the theory — find it cheaper
to hire a licence occasionally from, say, their local garage than to buy one
from the authorities.

5.3.6 We have considered the practical details of such a scheme in order
to determine the form of licence that might lend itself to the organisation of
such a market. The principal difficulties are the high value of the licence,
which causes problems of theft, fraud and finance; and the cost of adminis-
tration allowing for an adequate return on the high risk borne by dealers.
We have not found a satisfactory solution to these difficulties.

5.3.7 Feasibility of the daily licence. The daily licence is not so ambitious
as the differential annual licence. It need not be transferable, although
there is no reason why it should be non-transferable. The prices (of red,
blue and yellow licences) are fixed by the authorities. The licences are
numbered in large figures or letters to denote the day and are sold in two
kinds of book: (a) daily books consisting of a number of licences for a
given day intended for retailers (e.g. garages); (b) monthly or yearly books
consisting of one licence for each day of the month or year intended for
the use of individual vehicle owners. A rebate is given for licences not
detached from the book.

5.3.8 Unlike transferable licences, there are no purely technical difficulties
in a system of daily licences. It is simple and should not be difficult to
implement. In effect, the daily licence has much in common with the parking
tax. It has the considerable advantage, however, of applying to all traffic
entering the congested area, whereas the parking tax misses through-traffic
and other non-parking traffic. On the other hand, while the parking tax
can be gradually reduced towards the outer regions of the area, the daily
licence involves an abrupt change in tax at the zone boundaries. Tentative
attempts to delineate zones in particular towns have shown that in some
towns it is not easy to find a ‘ natural ’ boundary. These sharply discrimina-
ting boundaries inevitably result in inequity. The division of areas into
three grades of zones is an attempt to ease the discrimination but clearly it
can only be a partial solution to this difficulty, while introducing undesirable
complications into the system.

5.3.9 In addition, the daily licence shares much of the arbitrariness of the
parking tax. The short journey pays the same as the long journey. Long-
distance traffic with a red licence would be encouraged to pass through all
zones where congestion had been successfully reduced, thus diminishing the
apparent fairness and reasonableness of the system.

5.3.10 Economic advantage. The economic effects of a daily licence
system would be similar to those of a parking tax. The main difference
would be that non-parking traffic would have to pay. This is a considerable
advantage for the daily licence over the parking tax. But, as already noted,
the effect on through-traffic might still not be very satisfactory, because all
long-distance traffic arriving at one side of a red area, but having already
obtained a licence for use in another red area, would not be deterred from

Printed image digitised by the University of Southampton Library Digitisation Unit

crossing through the area, but rather would be encouraged to do so by the
easier conditions prevailing as a result of the scheme. The problem of
through-traffic is therefore only partially solved.

5.3.1 1 Another advantage of the daily licence over the parking tax is that it
may be used specifically for restraining peak hour congestion. In some
other respects, however, it is inferior to the parking tax. We have pointed
out that a parking tax may be tapered off towards the edge of the congested
area, but with a daily licence there is no similar way of avoiding the abrupt
boundary, and economic inefficiencies must result from the sudden, arbitrary
change in price from one side of the boundary to the other.

3.3.12 The fixing of prices presents another difficulty. Unlike the parking
tax, which can vary from town to town, it would not be practicable for a
large number of towns to issue their own licences. If the system were
to be widespread, the same licences, with the same prices, would need to be
valid throughout the country. This would mean that the best available
price in some places would not he satisfactory.

5.3.13 Clearly the relative weight of these various considerations varies
from one area to another. On balance, the advantage of including non-
parking traffic in its coverage seems to more than offset the economic dis-
advantages of the daily licence vis-a-vis the parking tax. If a nation-wide
system of daily licences were adopted there would be nothing in the system
to prevent parking taxes also being used in some places.

Other Indirect Congestion Taxes

5.4.1 Taxes on vehicle-ownership, petrol and parking are not the only
ways of indirectly taxing ro ad users, but we do not think that there are any other
such taxes which by themselves, and within the hmits of what is practicable
and relatively harmless to other activities, can produce much benefit on the
roads, although they may prove useful in combination with other methods.

5.4.2 One such measure is the poll tax on employees in congested areas.
The main arguments in favour of a poll tax do not concern the use of the
roads, and the more important aspects of such a tax lie outside the scope of
this report. We are concerned only with its likely effect upon road traffic.
The volume of road traffic would only be affected inasmuch as the tax caused
changes in the use of property, assuming that new construction was under
strict planning control. The demand for road space would then be affected
mainly as follows: —

(a) Organisations with large numbers of relatively low-earning employees
would tend to move out and their places taken by organisations with
fewer, hut higher-earning, employees. The number of car owners
among the latter, and their desire to travel by car, might be as high as
among the former, thus giving no relief on the roads, except possibly
on the buses. It is conjectural what the overall net effect of such
changes in property use would be.

(b) Large employers of labour would tend to be replaced by organisations
which generated less commuter traffic but more commercial, business

Printed image digitised by the University of Southampton Library Digitisation Unit

or tourist trafi&c. Without much more information on land use and
its effect on trafhc requirements, we cannot assess the net result of
these changes.

(c) A poll tax would encourage further mechanisation of labour, such as
lift operating, book-keeping, dictation and retail service. The level
of car ownership among the classes of labour affected is low.

(d) In the long run a poll tax would lower the demand for building new
offices. Whether this would affect the actual amount of new building
would depend largely upon the effectiveness of planning controls as
an alternative restraint. To some extent a poll tax and planning
controls are substitutes. Also in the long run the growth of car-
ownership among the lower income-groups will weaken some of the
above arguments against the poll tax.

5.4.3 Supposing that the effect of a poll tax was to remove an appreciable
volume of the existing commuter traffic from the roads, we believe that this
traffic would soon be replaced by other commuters. The level of commuter
traffic is limited now by heavy congestion and parking restriction. There
is ample evidence from statistical studies of car ownership and traffic speed
in towns that without some additional constraint which is not provided by
a poll tax, any reduction in congestion would be taken advantage of by the
reserve of car-owning commuters.

5.4.4 On balance, therefore, without undertaking the long and complex
study which a full examination of a poll tax implies, we have little doubt
that its effect upon the volume of road traffic would be of minor value.

5.4.5 Other taxes directed at city centres, such as a property tax, would
doubtless reduce the demand for new building. It is questionable whether
the volume of new building would be affected unless the tax were very high.
If it were not, it could do little to reduce the amount of traffic on the roads.
We know of no other similar tax which would have even as much effect on
traffic as a poll tax.

6. THE FEASIBILITY OF DIRECT CHARGING

6.1.1 Having considered methods of charging indirectly for road use,
that is, charging for some other product or service more or less closely
correlated with road use, we now turn to the possibilities of charging directly
for road use itself. Toll-gates have, of course, long been used in many
places and are still used on bridges and tunnels and on foreign motorways
with few points of access. But for ordinary roads in urban areas they are
costly and inefficient and impede the flow of traffic, and even with modern
refinements we do not regard them as practicable.

6.1.2 In our search for ideas we have spread the net widely in order not to
exclude any radical new methods which, though seemingly fanciful at first
sight, might eventually become an appropriate feature of a motorised com-
munity. With modern techniques many methods of charging are technically
possible, and in comparing them we have had to take account of their cost,

Printed image digitised by the University of Southampton Library Digitisation Unit

as well as their closeness in meeting the requirements specified in Section 3.
The various methods which have been propo&ed fall into two basic
categories ; —

(a) Off-vehicle recording systems, whereby the charges payable by vehicles
are recorded elsewhere than on the vehicles. These systems are
analogous to telephone charging methods.

(b) Vehicle metering systems, whereby the charges are registered on the
vehicles themselves. These systems are analogous to taxi-meter
methods of charging. They may be driver-operated or automatic.

Off-vehicle Recording Systems

6.2.1 In off-vehicle recording systems every vehicle is fitted with a piece of
equipment which can be automatically identified, and a record made, by
detectors placed in, over or beside the road at suitable pricing points. The
recording that a vehicle has passed a certain pricing point could then, in
theory, be used in one of two ways. The first, which we call point pricing,
involves setting up pricing points within congested areas, vehicles being
debited with the appropriate charge when passing any pricing point. The
alternative is a type of continuous pricing which involves setting up pricing
points on the borders of congested areas, and charging the vehicle according
to the time spent in the zone, as deduced from recordings at the points of
entry and exit.

6.2.2 Any continuous pricing system of this type would have to enable the
charging to cease when vehicles enter private premises, and to be reduced
in some parking places. This raises the considerable technical difficulty of
monitoring private entrances and parking places within pricing zones. This
cannot be done at an acceptable cost. Continuous pricing systems of this
kind would also require equipment of a higher standard of accuracy than
that required for point pricing. This is because under a point pricing system
occasional errors could be tolerated; no one would mind very much if a
vehicle passed 50 points and was charged for only 49 of them. But under
continuous pricing little errors could have serious results. If a vehicle was
recorded as entering a zone and its corresponding record of exit was missed
or mislaid, it would find itself being charged for an indefinite period. This
sort of error could not be tolerated, and methods of avoiding it would add
considerably to the cost of the equipment. For these technical reasons we
have concluded that any feasible off-vehicle system would work only on a
point pricing basis.

6.2.3 Off-vehicle systems comprise the following stages: —

(a) Identification of vehicles at the pricing points.

(b) Transmission of information from the pricing points to a central
computing station.

(d) Collection of payment.

6.2.4 Vehicle identification. Vehicles must carry identification units which
enable their presence to be recorded by roadside apparatus. The recording

Printed image digitised by the University of Southampton Library Digitisation Unit

must be in a suitable form to comprise the input data of the computing
equipment. The system must be capable of distinguishing between, say,
30 million different vehicle identities. Although on general grounds this
problem appears soluble, none of the methods proposed to us have come
near the cost levels we set in paragraph 3.4. We have enquired about
optical, electro-magnetic, radar and sonic methods, and the only serious
proposal put to us was the electro-magnetic Link Tracer suggested by
Professor William Vickrey for vehicle identification in Washington, D.C.C"*)
The capital cost quoted for the vehicle, roadside and computing equipment
in Washington was £12 10s. Od. per vehicle, but we have been advised that
for a population of 30 naillion vehicles this cost could easily double. How-
ever, even £12 10s. Od. is a good deal higher than the £5 per vehicle that we
allowed.

6.2.5 Transmission of information. The simplest method would be to
accumulate the records at each pricing point in the form of paper or magnetic
tape and to transport the tapes to the computing station at regular intervals
by van. The alternative would be to transmit the information automatically
by telephone or by radio signals. Separate transmission channels could be
provided for each pricing point but this could be wasteful. A more economic
arrangement might be for a number of pricing points to share a transmission
channel, and for temporary storage capacity to be provided at the pricing
points to prevent the loss of records generated simultaneously at several
places sharing the same channel. Temporary storage capacity at the com-
puting station input would also be needed to prevent loss when a large number
of channels transmit simultaneously. The transmission of data from the
pricing points to central computing stations does not appear to present any
special difficulty.

6.2.6 Processing of data. This also does not appear to be difficult with
conventional equipment. All charges would have to be sorted by vehicle
identification number, added and printed out at intervals. Local accounts
could be retained for mailing, and other accounts (the region being indicated
by the number) sent to their respective regions either direct or through a
central clearing house. As a final stage charges would have to be multiplied
by the appropriate charging factor (which might be 1 for a private car, 2
for a lorry, | for a motorcycle, etc.). Bills could be sent at monthly or
quarterly intervals as required.

6.2.7 Collection of payment. The difficulties and costs of bill collection
from motorists would need careful study. If there were 10 million accounts
and if each year one account in a thousand failed to pay after the usual
reminders, there would be some 10,000 cases a year to follow up, some of
them through the couits. The authorities could obtain some protection by
making renewal of vehicle licences conditional on all road charges being
paid up to date. Nevertheless, many difficult cases would remain, and it
would clearly be undesirable to allow large numbers of people to become
debtors for the considerable amounts involved in road pricing. One solution
might be to require payment in advance. The equipment in the computing
station could be programmed to print out the vehicle numbers of those falling
into debt, and the owners of the vehicles involved could be dealt with
immediately.

Printed image digitised by the University of Southampton Library Digitisation Unit

6.2.8 Conclusions. The advantages of a fully automatic system, with, ample
price flexibility, in which the vehicle-owner has no more to do than periodically
send a cheque, are obvious. But the above brief sketch of off-vehicle
systems indicates a number of disadvantages in comparison with automatic
metering systems :

(a) They require equipment for the storage, transmission and processing
of data. Having regard to the large number of places where vehicle
movements would need to be recorded, and to the present cost of
standard equipment, the provision of this equipment — while not
difficult — ^would undoubtedly be expensive.

(b) They require centralised book-keeping, invoicing and progressing of
bills. Payment in advance would not always be easy to obtain.

(c) They threaten the privacy of vehicle users, insofar as they enable
vehicle movements to be traced by public authorities. (This dis-
advantage would be mitigated by the greater ease with which the
police would be able to trace stolen cars, and by the usefulness to
some vehicle owners — and to road planners — of vehicle movement
records).

6.2.9 In view of these disadvantages — which are not sufficiently offset by
advantages — ^it appears to us that off-vehicle systems should be considered
only if they offer considerable financial savings over meter systems. In fact,
all the systems suggested to us cost much more than the meter systems which
we are about to describe, and we therefore conclude that, on the basis of
present information, off-vehicle systems should be rejected in favour of
vehicle metering systems.

Driver-operated Meters

6.3.1 These systems work on a continuous charging basis. Zones are
defined in congested areas and are allocated colours which indicate the
prices to be charged within the zones. Thus a conurbation might be divided
into three zones, an expensive purple zone in the centre, encircled by a
cheaper green zone in the inner ring, encircled again by a still cheaper pink
zone in the suburbs. Outside the pink zone the roads are free, subject of
course to any fuel tax in force. In other cities two zones might be sufficient,
and in smaller towns there might be only one. The colours of the zones
are displayed electrically at all points of entry and exit and can be switched
so that during off-peak hours during the day the purple zone can be derated
to green and the green zone to pink, and the pink zone can either be derated
to a fourth colour, say brown, or dezoned altogether. During the evening
and at night, and on Sundays and Bank Holidays, all zone signs might be
switched off and the zones made free.

6.3.2 Every vehicle entering a charging zone must carry a road meter.
When the meter is switched on and working it shows a coloured light, indicat-
ing the rate at which it is running. Thus the driver switches to pink when
he enters a pink zone and if he later drives out of that zone into a free zone or
a purple zone he switches his meter accordingly. The meter is fixed at the
bottom centre on the inside of the windscreen within comfortable reach of
the driver’s left hand, so that the switching operation would be comparable

Printed image digitised by the University of Southampton Library Digitisation Unit

with that of switching a trafficator. Thus a driver might need to make
three such operations on a journey from the outskirts of a conurbation to
the centre, which is probably a small fraction of the number of driving
operations he would be saved by driving in the less congested streets.

6.3.3 If pink represents a charging rate of Jd. perminute, the meter is designed
to register or consume at that rate when switched to pink, and similarly for
the other colours. One or two lower rates for parking may be added, thus
providing what is in effect a personal parking meter and removing the need
for pavement meters.

6.3.4 Different classes of vehicle are issued with different classes of meter,
so that if a private car, say, carries a Class A meter, a 10-ton lorry might
carry, say, a Class D meter with a higher charging rate.

6.3.5 We have examined two possible systems of driver-operated meters
and have submitted details of them to the Ministry of Transport. Here
we shall confine ourselves to a simple description of each of them in turn.

6.3.6 A clockwork meter

6.3.6. 1 Timing mechanism. This consists of a spring-driven motor
working through a clockhke mechanism. The five colours are selected
by means of a gearbox. The spring is electrically wound and is connected
with the vehicle’s electrical circuit.

6.3.6.2 Through its gears, the mechanism moves a counter which is
visible to the driver. The meter is initially supphed with a number of
shillingsworth (e.g. 100) set on the counter, representing the value of time
being purchased. As the meter is used, the counter runs down at the
predetermined rates, and when it reaches, say, ten shillings below zero an
expiry flag appears together with an indicator fight. The purpose of
allowing the meter to go below zero is to give a period of grace in which
to exchange the meter for a fresh one.

6.3.6.S Method of payment. The motorist purchases a meter at a post
office or authorised depot. The value of time set on the meter may vary
widely according to customers’ requirements. When the meter is expired,
it can be exchanged for one that has been already reset. At the same time
any excess units used, i.e. units below zero, must be paid for. In view of
the workshop facilities that garages possess and the unfair temptation
that would be put in their way, the authorities might not wish to use them
or any other agencies, apart from post offices and official meter depots,
as distributive outlets. The resetting of meters would be carried out at
depots by insertion of a special key, this action involving the destruction
of a lead seal or similar security device, which would afterwards be replaced
by the depot.

6.3.6.4 Other features: (a) Visibility. The zone indication is provided by
coloured illuminated translucent sectors shown through a window in the
face of the' meter. On the side facing the driver there is another window
with a coloured disc showing.

Printed image digitised by the University of Southampton Library Digitisation Unit

(6) Appearance. The meter is box-shaped, about 4 in. X 3 in. X 3. in.
On the front is the lighted window. On the top is a place for replacing
bulbs. On the side facing the driver is the selector knob, the counter
window and the window showing him the zone colour. On the bottom is
the connection with the vehicle battery circuit and the fixing bracket.
The meter cannot be fitted permanently, since it must be exchanged
periodically; at the same time it must not present too easy a prey for
thieves.

6.3. 6.5 Security against fraud. The final casing may be an aluminium
die casting with provision for sealing over a sunken screw. The resetting
of the meter would then require a special key. These arrangements do
not appear sufficient by themselves to prevent fraud, and further investiga-
tion is required on this point.

6.3.6.6 Cost. A preliminary estimate, assuming large quantities, indicates
a cost of £8-£10 per meter. It is hoped that closer attention to production
engineering requirements will bring the cost down. The administrative
costs of the scheme would arise mainly from the running of the meter
depots.

6.3.7 An electrolytic meter

6.3.7.1 Timing mechanism. A ‘throw-away’ electrolytic timer is used.
Cylindrical in shape, it is inserted into the meter by the driver and, when
pushed home, forms part of a circuit connecting the meter lamp with the
vehicle battery. Also included is a network of resistors so that the timer
runs down at the correct predetermined rates when switched to different
colours. An essential feature of the timer is that its hfe is accurately
gauged. When it reaches its designed end-point the circuit is broken and
the meter lamp cuts out. Before this happens, when the timer is down
to, say, its last ten shilhngsworth, a warning light appears, to remind the
driver that his timer will soon need replacing. When it is finally exhausted
it can be removed from the meter and replaced.

6.3.7.2 The timer can pass an accurately predetermined electrical charge
before exhibiting an electrical change sufficient to control a switching
circuit. No basic difficulty is anticipated in producing a timer which
wfil perform this function satisfactorily. Several different designs have
been discussed, and similar devices are actually in use for various industrial
purposes in the United States. The question that remains is whether a
satisfactory design can be produced cheaply enough for this purpose;
and the answer to this question requires investigation.

6. 3.7.3 Method of payment. The main reason for preferring electrolytic
timing to a more conventional mechanical method is that it overcomes
difficulties of payment. Timers are bought “ over the counter ” from
garages or post offices in units of, say, £1, £2 or £5. Thus the meter need
never be read or removed for resetting. This is convenient for both the
motorist and the authorities. The revenue due to the Government could
be collected as an excise duty from the timer manufacturers.

Printed image digitised by the University of Southampton Library Digitisation Unit

63.1 A Other features: (d) Visibility. The zone colours are displayed
by a bulb about as bright as a sidelight beamed through a sector of coloured
filters. The colours must be distinct at a sufficient distance to facilitate
supervision but need not be as brightly illuminated as a trafficator.
Hence, although the pricing period would include some hours of darkness
in the winter, no dazzle is expected from the meter lamp. The coloured
light can also be seen from the rear.

(i) Initial flasher. When the meter is switched on or turned from one
colour to another, the bulb flashes for 20 seconds before stabilising.

(c) Parking. If the meter is required for parking, a clock dial may be
fitted to the face of the meter and used as a parking disc. A time switch
can also be fitted for the benefit of drivers who park during charging
hours and want their meter automatically to stop charging when parking
becomes free.

{d) Appearance. The design envisaged is circular, about inches in
diameter and IJ inches deep. The face of the meter contains a window
for the lamp. On top is a place for replacing bulbs and fuses and an
aperture for the insertion and extraction of the timer: and underneath is
the lead (preferably encased in a bracket) to the vehicle’s battery circuit.
On the rear of the meter, the side facing the driver, is a selector knob with
six positions suitably marked, and a rear window aperture. When required,
a lock can also be incorporated to guard against theft of the timer; this is
especially desirable on open cars and motorcycles. Every meter also
bears an identification number.

6.3.7.5 Security against fraud. Those parts of the meter which could
profitably be tampered with are potted in plastic and wholly encased.
The timer chamber and the circuit are so designed that only an accurate
forgery of the timer, both of its shape and its electrical properties, corild
work. Spot inspection of both meter and timer can be made by the
police in the same way that licences are checked.

6.3.7.6 Since there is no need ever to remove the meter from the vehicle,
there is no difficulty in maintaining control over meters issued. When
a meter is issued it may be registered in the log book of the vehicle for
which it is intended. An arrangement is desirable whereby the meter is
returned when the vehicle is scrapped. Having taken these steps to
ensme as far as possible that only one meter exists for each vehicle, a meter
test can be included in all regulation vehicle tests.

6.3.7.7 Meter ‘ dodging ’ — switching off when there is no enforcement
officer in sight — ^is discouraged by the tell-tale 20-second flasher.

6.3.7.S Cost. It is thought that the meter could be produced (excluding
the parking time switch) for a price of about 30s. The principal objective
of further development is to produce a timer for less than a shilling. The
production cost wiU probably be much the same for a £1 unit as for a
£5 unit; the proportion of revenue lost being rather high on the former
and very low on the latter, so that some incentive may be desirable to
promote the sale of the larger unit.

Printed image digitised by the University of Southampton Library Digitisation Unit

Automatic Meters

6.4.1 The principal drawback of the driver-operated meter is that it
adds to the responsibilities of both driver and traffic authority, although
of course this is expected to be more than offset by the easier traffic con-
ditions that result. The automatic meter tries to eliminate much of this
responsibility by placing control apparatus in the road which switches meters
on passing vehicles to the correct position. We have examined four possible
systems of automatic meter, details of which have been submitted to the
Ministry of Transport. Two of them are similar to the driver-operated
system with electrolytic timing mechanism except that the switching operations
are carried out automatically. These are referred to as continuous systems
A and B. The other two systems work on a point pricing basis and are
referred to as point systems A and B.

6.4.2 Continuous system A

6.4.2. 1 Under this system the method of charging and the arrangements
for payment are the same as for the driver-operated meter described in
6.3.7 above. The main difference between the two systems is that, whereas
in the latter the motorist sets his meter to the appropriate running rate,
under the automatic system this function is performed for him by an
additional switching circuit which operates in response to signals received
from road-sited transmitters installed at the zone entry and exit points
and at intermediate points within the zones. Manual control of the auto-
matic meter is still retained for the selection of parking rates. A manual
over-ride is also provided to enable a motorist to stop his meter charging
when he enters private premises situated within a charging zone. The
over-ride is itself over-ridden, however, if subsequent to its use the meter
receives a signal from a transmitter to levy a charge.

6.4.2.2 Another important difference is that only one colour indication
is needed.

6.4.2.3 Timing mechanism. This is exactly the same as in the driver-
operated electrolytic system.

6.4.2.4 Switching mechanism: (a) Road Equipment. This consists of a
short-range low frequency transmitter which can be set to transmit a
signal at one of several possible frequencies, according to what charge is
required at the time. The signal is radiated from a cable laid across the
whole width of the road and is picked up by a vehicle passing in either
direction. The meter on the vehicle is automatically switched to the
charging position corresponding to the signal frequency. The outer-
most cables around a charging area will carry the OFF signal, so that
vehicles entering from the free area outside wiU be unaffected by the
first cable they cross. But a Uttle further on they will cross another cable
which will set their meters to the required charging position. When
travelling inside the area they may cross a number of cables, some of them

Printed image digitised by the University of Southampton Library Digitisation Unit

mternal controls which will not affect meters which are already set to the
correct rate, and some denoting entry to another zone with a different
rate. Quite small areas, or even single streets, could he made separate
zones. When charging ceases each day, aU cables are automatically reset
to transmit the OFF signal.

The transmitters are very compact and the little power they require
should normally be obtainable from the street hghting system. In fact,
the most convenient place to install the transmitters, which could then be
housed in sealed cast-iron containers, might be to attach them to the base
of existing light standards. Inside each container there would also be an
electrically operated perpetual time switch which would control the daily
and weekly sequence of frequencies to be transmitted. It would be
possible for the police or traffic wardens to change the normal frequency
sequence on special occasions such as bank hohdays.

(b) Vehicle equipment. The vehicle equipment is so designed that the
switching arrangements inside it will operate in response to the frequencies
picked up from the cables in the road. On receipt of a signal to charge
at a certain rate the meter will charge at that rate until it receives a further
signal either to charge at a different rate or to switch off or until the manual
over-ride is used to switch off the meter. The signal is picked up by a
con attached to the underside of the vehicle chassis.

6.4.2.5 Method of payment. This also is the same as for the driver-
operated meter.

6.4.2.6 Other features: (a) Visibility. As in the case of the driver-
operated meter, indication that the meter is working is provided by a
light which is visible to an outside observer. With the automatic system,
though, the same light can be used for all the running rates. There is no
need to indicate that the meter is switched to the correct rate.

(Z>) Parking. The arrangements are the same as for the manual meter.

(c) Appearance. The location of the automatic meter is the same as
for the manual meter, i.e. at the inside centre of the windscreen, and the
only external difference in the appearance of the two meters is that in the
case of the automatic one no switch positions are needed for the running
rates.

6.4.2.V Security against fraud: (a) Meters. Whenever a meter undergoes
a change of rate, the meter light will blink OFF and ON for a fraction of
a second. A meter which has been tampered with in such a way that it
no longer responds to the normal automatic switching signals, e.g. one
whose coil has been screened, will fail to show this indication and can
thus be detected quite easily. As in the case of the manual meter the
circuit will be ‘ potted ’ and therefore completely inaccessible.

(6) Timers. Production of a counterfeit timer would be comparable
in difficulty to forging a bank note. If necessary the police could be given
authority to make spot checks of timers on a small sample basis.

6.4.2.S Cost. At the moment it is envisaged that the cost per unit of
the vehicle equipment would be about £3, excluding fitting costs which

Printed image digitised by the University of Southampton Library Digitisation Unit

would be small. The road equipment would probably cost about £50
per point, excluding the cost of connection to the mains and the cost of
laying the road cables. This latter operation, however, could be carried
out without actually digging up the road; aU that is needed is a slot about
an inch wide and few inches deep, and special drilling apparatus could
probably be made to cut the slots quickly and cheaply.

6.4.3 Continuous system B

6.4.3.1 This scheme is based on coloured charging zones with an option
of three charging rates denoted by three colours, used in exactly the same
way as in the driver-operated systems already described.

6.4.3.2 The timing mechanism. Electrolytic timers are used as in the
A system, the main dilference being that in the B system the timer is not a
“ throw-away ” item but is used repeatedly in the same meter. A constant
current is made to flow through a sealed electrolytic cell depositing copper
from one electrode to another and, when this process nears completion,
say after one hour at the highest rate, the current is automatically reversed
and the process is repeated in the opposite direction, and so on. Every
time the cycle is completed an electro-magnetic counter is moved down
by one unit. The counter, which on issue might register, say, 20 units,
would eventually reach zero and would automatically switch out the
circuit.

6.4.3.3 The switching mechanism. This is much the same as in the
A system.

6.4.3.4 Method of payment. When the counter needs resetting, a small
cassette containing the electrolytic timers and the counter is removed from
the meter and taken to an appropriate centre. The cassette must be
handed in while there is still a refundable amount showing on the counter,
and either reset on the spot or exchanged for one already reset; or
alternatively a spare cassette can be kept ready and inserted as soon as
the old cassette expires.

6.4.3. 5 Other features: (a) Visibility. Four lamps are employed, one
colour for each charging rate and a fourth no-charge lamp to indicate that
the equipment is switched on. This last lamp will be visible both from
inside, at low intensity, and outside the vehicle.

(b) Parking. Two alternative systems may readily be accommodated.
A flat rate for parking could be used, set somewhat lower than the lowest
running rate. Alternatively, the ignition circuit could be carried through
the equipment and would cause a lower rate of charge to be levied when
the ignition was switched off than when the vehicle was moving within
that zone.

(c) Appearance. The meter is box-like in shape, 3i in. X 2-J- in. X If in.
In the front are four small light apertures. The no-charge light is also
visible from the rear. There is also a window showing the number of
units on the counter.

Printed image digitised by the University of Southampton Library Digitisation Unit

6A.3.6 Security against fraud. PoKce or traffic wardens would be
required to observe the light indications which must be shown in the
appropriate zones. These indications should be readily apparent from
the kerb-side, even from quickly moving vehicles. However, it might be
possible to adjust either the indicator or the count in the meter if the seal
were broken, and it is therefore suggested that portable transmitters
producing changes in the lamp indicators could easily be provided for
officers to detect any cheating.

6.4.3. 7 Cost. It is thought that the equipment to be mounted on the
vehicle, i.e. the pick-up coil and meter, could be quantity produced for
about £4-£5. The zone-switching signal transmitters could be produced
for about 25s. each. The operating costs would arise mainly from the
cost of running meter depots for handling cassettes.

6.4.4 In the point pricing systems the meter does not run continuously
while in a charging zone but registers one or more units every time it passes
a charging point. It is in effect a sophisticated toll-gate. The advantage
of point pricing is that the price may be varied from point to point and the
points themselves may be deployed more or less densely, so that highly
flexible pricing is possible. The disadvantages are that, unless the points
are spread in very large numbers, they can be to some extent avoided, and the
fewer there are the less accurately do they relate to costs and the more
arbitrary and anomalous are their effects upon traffic. But large numbers
of points add to the cost.

6.4.5 Point system A

6.4.5.1 The counting mechanism. The meter is designed to be carried on
the outside of the vehicle, either near the bumper or on one wing. It is
issued containing a definite number of units. Whenever the vehicle passes
a charging point, a number of units, between one and ten, is deducted.
In its basic form the counter produces no visible sign, so that it is not
possible to read the counter in any way. Provision is made for part of
the unit to change its colour as soon as a predetermined number of impulses
is recorded, and provision can also be made for a second change of colour
to appear at any intermediate stage. Thus the counter could be issued in,
say, green form when unused, turning to yellow when almost exhausted
and red when completely exhausted. A visual indicator (needing some
training to read) can be attached at an extra charge, and a decimal con-
verter can be added at a further charge.

6.4.5.2 Power requirements for the meter are very small and may come
from one of three sources; (a) an independent internal power supply
generated by the movement of the vehicle; (b) a self-contained long life
battery; (c) the vehicle battery.

6.4.5.3 The switching mechanism. Electric cables carrying a very low
current are laid transversely across the road, either in or on the surface,
and are energised continuously so that any vehicle moving across a cable
receives an impulse. Vehicles stopping on top of a cable do not pick up
more than one impulse. It is proposed that the cables be laid in groups

Printed image digitised by the University of Southampton Library Digitisation Unit

of up to ten at each charging point, so that the number of impulses to
activate meters at each point can be varied accordingly. Changes of
charging rate could be made manually at each charging point or auto-
matically from central control stations.

6.4.5.4 Method of payment. Two methods are suggested. A prepay-
ment meter may be sold with a given capacity and exchanged when
exhausted. Since this meter must be removed from the vehicle periodically
it cannot be fixed permanently and must therefore be securely locked to
the vehicle. Alternatively, a credit meter can be issued which can be
fixed permanently to the vehicle and taken at intervals to authorised
stations to be read and paid for. This type of meter would be registered
in the name of the owner. Precautions against theft need further investi-
gation.

6.4.5. 5 Other features: (a) Visibility. No illumination is provided. As
long as the meter is in working order evasion is not possible except by
carrying an expired prepayment meter, in which case the chances of
avoiding detection need be no higher than with an expired vehicle ficence
at present. Illumination may therefore be unnecessary.

(h) Parking. Cables are laid along parking places and impulses emitted
at fairly long intervals corresponding with the required parking rate.

(c) Appearance. The meter is envisaged as an encapsulated box about
6 in. X 3 in. X 1 in. The only feature on the surface is the window
showing the colour and the counter.

6.4.5. 6 Security against fraud. Tampering with the meter itself is
considered very unlikely because of the degree of encapsulation possible
and the complexity of the mechanism. Screening, however, may be
possible and investigation is required on this point.

6.4.5.7 Another danger of fraud seems to lie in the possibility of putting
a prepayment meter out of action, although the manufacturers claim that
the meter will be almost indestructible. The best way of defeating the
prolonged use of out-of-action meters may be to change the style and
colour of the casing at frequent intervals.

6.4.5.S Cost. The tentative cost estimates are as follows :

For the tatic meter : £3 to £5
For the binary indicator : £3 to £5
For the decimal converter; £3 to £5.

We have suggested that the static meter alone would be sufficient for a
prepayment unit, but a credit meter might require the two attachments.

6.4.5.9 The cost of ground equipment will depend upon the number of
charging points adopted. For a national system covering all congested
areas we estimate that the minimum number of points needed to give
tolerably efficient results would be about 20,000. The maximum cost
per point is estimated at £250, but may well prove to be lower than this.
Thus we may set a maximum cost for ground equipment at £5 million.

Printed image digitised by the University of Southampton Library Digitisation Unit

6.4.5.10 The operating costs of the system consist mainly of the costs of
the meter depots. The overall cost of maintaining and administering a
credit meter should be comparable with that of a gas meter. No estimate
is available of the costs of running an exchange scheme for prepayment
meters.

6.4.6 Point system B

6.4.6. 1 The counting mechanism. Every vehicle would carry a unit
containing a pick-up device, a receiver and an electro-magnetic unit
counter. The unit would be on the outside of the vehicle and wired
permanently to the vehicle battery. Every time the vehicle passed a
charging point (consisting of a loop in the roadway) the receiver would
pass one or more pulses through a transistorised circuit which would
advance the counter by the same number of units.

6.4.6.2 The switching mechanism. Three arrangements could be used
together. Long loops could be placed in the road to transmit pulses at
a rate corresponding to the charging scale in use. The frequency proposed
is below 10 kilocycles so as to avoid interference from other communica-
tional and navigational signals. The loops would be laid below the
surface extending for the full width of the road and from 30 feet up to
a city block in length, although this would cause difficulties for road
repairs. The current used would be of the order of 1 amp. To provide
for differential charges the pulse rate could be changed by time switch.
Alternatively, short loops could be placed at intervals on the road trans-
mitting either pulses or a continuous wave. To provide for differential
charges some of the loops could be switched off separately by time-switch.
In all three arrangements small transmitters would be used, mains-operated
and conveniently mounted on lamp posts, walls, or beneath the pavement.

6.4.6.S Method of payment. It would be necessary for the meter to be
read periodically and it is suggested that this could be done at garages.
The vehicle owner would be under an obligation to have his meter read
at least once every three months, say. Any approved garage could do
this and accept payment, in return for a payment voucher which would
bear the date and which could be fixed to the windscreen. The payment
would then be remitted to the local authority who would keep an account
for each meter issued. If a meter fell overdue for payment there would
thus be two checks, the payment voucher and the account. The authority
could refuse to renew the licence of any vehicle with an overdue account.

6.4.6.4 It is proposed that the issue of payment vouchers should be
made by a locked and sealed machine installed at each approved garage
and operated by a garage attendant. The machine ‘ reads ’ the necessary
information on the previous voucher (inserted for the purpose) and issues
a new voucher in accordance with this and the setting of two sets of dials
by the attendant. It can refuse to accept a setting that leaves the motorist
in debt. The purpose of the payment machine is both to make payment
fast and simple, and to prevent fraud by either motorist or garage.

Printed image digitised by the University of Southampton Library Digitisation Unit

6.4.6.5 Other features: (a) Visibility. A small lamp would be situated
inside the vehicle where it could be seen by the driver and observed by a
warden as the vehicle passed. When the vehicle meter picked up a pulse
or signal, the lamp would flash, thus informing the driver that he was
being charged and at the same time providing an outward indication that
the meter was functioning.

(h) Parking. The system is readily adaptable to use on parking sites.

(c) Appearance. The best location for the meter has not yet been
decided but it would be placed on the outside of the vehicle. It would
consist simply of a hermetically sealed unit, with a single glass or perspex
window provided for the counter, and with leads to the battery and
indicator lamp. The vehicle registration number might be reproduced
on the inside of the counter window. The colour of the case could denote
the price class of the vehicle.

5.4.6.6 Security against fraud. As already stated, the meter would
consist of a hermetically sealed unit and the vehicle registration number
could be inscribed inside the counter window on issue. The range of the
counter would exceed the maximum requirement of any vehicle so that
the meter need not normally be removed during the hfe of the vehicle.
Meters should be obtainable only through the authorities and should be
returned when the vehicle is scrapped. The counter would be readily
observable when the vehicle was stationary and the current payment
voucher would need to be displayed, probably on the windscreen, so that
wardens could verify that payment was up to date. Since the meter would
be on the outside of the vehicle, any attempt to screen the pick-up device
would be readily detectable. The wiring between the unit and the vehicle
battery is the most difficult problem: a simple arrangement is favoured,
which could be easily inspected, consisting of a flying lead hard-soldered
and sealed to the battery cable immediately adjacent to the battery terminal.

6.4.6.7 The indicator lamp would be an effective way of ensuring that
meters were properly functioning. Vehicles could be observed passing
any charging point. Officers could be provided with test torches which,
when directed at a vehicle, would cause the lamp to flash and the counter
to register one unit. (The cost of one unit under this system would only
be about one-tenth of a penny so this need not cause any difficulty).

6.4.6. 8 Cost. In full production the cost of the vehicle equipment is
tentatively quoted as £5-£10. Of this cost 65 per cent is attributable to
the electro-mechanical counter, so the final cost would depend largely on
whether this could be reduced. The road equipment is estimated at
£100-£150 per unit (excluding installation costs); the number of units
needed for a nation-wide scheme being possibly 20-30,000, excluding
parking sites. The cost of the payment machines, of which 10-20,000
might be needed, is not known.

Zone Marking

6.5.1 If an automatic system is adopted, zone marking presents no problem,

since it exists only to inform, not to instruct. With a driver-operated meter

Printed image digitised by the University of Southampton Library Digitisation Unit

it must be unmistakable. It must teU the driver which rate is required and
it should permit changes of rate to be made at certain predetermined times.
If rate changes are to be made regularly and quickly, electrical signs will be
required.

6.5.2 It is suggested that zone marking consists of primary marking at all
points of entry and exit, and secondary marking within the zone. The
primary marking might be achieved by beacons, not necessarily of the same
shape as crossing beacons. The lighted part of the beacon could consist
either of a large white glazed shade containing bulbs of the necessary colours,
or of separate lamps after the fashion of traffic lights. As it is important
that the system should be simple, it is unlikely that more than two colours
would often be required for the same beacon. The beacons would be
switched on at the correct time, either by hand or by time switch, and could
be switched from the off-peak colour to the peak colour, and so on, without
difficulty. The cost of a beacon has been estimated at about £42.

6.5.3 Secondary marking might consist of small sign-boards placed at
prominent places within the zone such as traffic hghts, parking sites, and
in other places where motorists might need to see them. These boards could
be fixed to lamp posts and other suitable posts already in existence including
the peripheral beacons. Their purpose would be to inform the public of
the general pricing regulations applying in that area.

6.5.4 Parking sites embraced by the scheme would need suitable notices
explaining the charging colour required and any time hmits in force.

Supervision and Enforcement

6.6.1 The operation of the price system (or any system of distribution)
requires supervision and enforcement. Most retail trades rely upon sales
assistants for supervision. Transport services, and many other service trades,
rely also upon inspectors. Many thousands of people are employed wholly
or partly on this function of supervision. Underpinning their authority
lies the power of the law, upon which the enforcement of the price system
ultimately depends. Both supervision and enforcement are necessary costs
which must be taken into account when discussing new systems of road
pricing.

6.6.2 One of the advantages of the present system of fuel tax and vehicle
hcences is that it requires very little supervision and enforcement. This
must be borne in mind when considering the advantages of other systems.

6.6.3 Where parking meters are used, their supervision is undertaken by
traffic wardens. In London it is estimated that the cost of wardens will
reach £600,000 a year when all parking zones are in operation. The
organisation needed to trace drivers who fail to pay excess dues and fines
adds a substantial further cost. Thus street parking systems are expensive
to enforce and supervise, but this is. partly because of the time limits.

6.6.4 With or without time hmits, parking meters require traffic wardens,
and their duties may seem naturally extendable to the supervision of road

Printed image digitised by the University of Southampton Library Digitisation Unit

meters, -which charge for both moving and parking. It seems unnecessary
that the routine inspection of road meters and the supervision of their
operation should be a function of the police.

6.6.5 With a driver-operated meter there are broadly two -ways of avoiding
payment: either by failing to operate the meter correctly or by tampering
■with its mechanism. With an automatic meter only the latter possibility
exists but there may be greater scope for exploiting it. With adequate
zone marking the amount of accidental avoidance should be small, and we
shall therefore confine our attention to deliberate avoidance by ‘ dodging ’
or tampering.

6.6.6 The dishonest driver’s willingness to commit fraud depends upon
the gain from the fraud, the probability of detection and the penalty. To
some extent the last two factors are substitutable, but the severity of the
penalty is in practice limited by public opinion. The gain from meter
tampering could be large-; it could give the beneficiary ‘ free ’ motoring for
an indefinite period. So the chance of escaping detection indefinitely must
be kept small. Heavy penalties would doubtless be permissible for this
more serious type of fraud. Experience of meter tampering in the gas and
electricity industries seems to suggest that, unless there is a coin box, this
sort of fraud is very rare; and we think that this is due to the conditions
surrounding the use of the meter rather than the inviolability of its design.
Provided that the fraud cannot be easily and permanently concealed, the
mere possibility of an occasional check wiU deter all but the criminal-minded.
If the meter can be retarded in some way and then rectified later on, the
checks must be random. Ignorance of the mechanism is another deterrent.
The registration of meters, making drivers accountable for them, is another
powerful deterrent. If drivers are permitted to obtain only one meter
per vehicle and are held responsible for it, they will hesitate to interfere with
it for fear of spoiling it. In short, given a reasonably secure design with a
confidential mechanism, combined with registration and a system of regular
and random checks, we would not anticipate much trouble with meter
tampering. These conditions give an advantage to the meters which can be
permanently encased and fixed to the vehicle.

6.6.7 Meter ‘ dodging ’ presents a different sort of problem because it
might not be regarded by many members of the pubhc as a criminal activity
and hea-vy penalties might therefore be impracticable. On the other hand,
the gain to be enjoyed from each act of ‘ dodging ’ would be trifling, and
only by frequent repetition could a substantial gain result. Given a meter
with good visibility characteristics, the chances of evading detection for
long would, we believe, be very small. The initial 20-second flasher suggested
for the driver-operated meter would be a useful feature.

6.6.8 A firm initial effort to obtain strict compliance might be advisable
because drivers with dishonest inclinations might hesitate to become the
‘ odd man out ’ by committing petty offences in fuU view of other drivers who
were observing the law. A high degree of compliance, once achieved,
might thus be self-perpetuating. This principle is used in Russia to obtain
payment on conductor-less buses.

Printed image digitised by the University of Southampton Library Digitisation Unit

6.6.9 It is impossible to predict how big an enforcement effort would be
necessary. But before weighing this effort against the advantages of better
road pricing one must take account of the savings that would accrue to the
police through reduced congestion. In Central London alone it is estimated
that the anticipated reduction in congestion would save about £800,000
a year in police costs.

Conclusions on the Feasibility of Direct Charging

6.7.1 We have now described the various ways known to us of charging
for the use of the roads. When charging for a commodity or service it is
logical to look first for a direct method of doing so and there are in fact
very few commodities or services which, if charged for at all, are not charged
for directly. The decision that the use of the roads by vehicles should be
charged for was made by Parliament in 1909. At that time some form of
meter would have been possible but not at a reasonable cost. Today,
half a century later, we can say the same thing about off-vehicle recording
systems; they are technically feasible but it is controversial whether they
are superior, or whether they would be acceptable to the public, and their
cost, for the moment, is high. On the other hand, it seems that a simple,
effective meter can now be produced quite cheaply with manual control,
and may be provided with automatic control for an extra cost which need
not be regarded as excessive.

6.7.2 None of the meter systems put forward has been developed to the
point where it could be adopted without qualification. Further development
work is needed which, although quite straightforward and not costly, will
require some encouragement from the Government. Given the signal to
proceed, manufacturers expect to produce fuUy developed prototypes and
detailed costing within a year.

6.7.3 After this period of development it should be possible to select the
most promising method or methods. In the meantime arrangements could
be made for operational and introductory trials. New methods must first
be tried out on a small scale and should, if possible, be introduced gradually.
Thus one would suggest a technical trial in a place where the issue of meters
to incoming vehicles is easily arranged, followed by an experimental scheme,
possibly in Central London. Any adjustment of prices, including fuel tax
and vehicle tax, could only be completed once a national scheme was in force.
During the introductory stage there would be anomalies. Relatively low
prices could be charged at first, accompanied possibly by local reductions
in vehicle tax.

7. THE PRICES

7.1 We stated above that, in general, road prices should equal the costs
arising from the passage of each vehicle. We also stated that these costs,
which are mainly congestion costs, vary from about 4d. a mile when the
average traffic speed is 20 mile/h to about 6s. a mile when it is only 8 mile/h.

Printed image digitised by the University of Southampton Library Digitisation Unit

i-UEL TAX

*3 5 ?

rnc^o 2.

^ ; a"

zi "^5 o

:^c=ii

_?3fTig £ :
ni r’fg = j

H o :
>> -’•
XX

CZrtff
^ Z o p

mr^-
HO
> m

ffl

o > >

'-'^TS

“r 7^ «o JT

PpT

<o 2"'2,^

5> °

-3

Printed image digitised by the University of Southampton Library Digitisation Unit

Methods of Charging for the Use of the Roads

This does not mean, of course, that these are the right prices to charge, because,
as soon as a price is introduced and causes the volume of traffic to fall, con-
gestion costs will also fall, and in much greater proportion. The prices for
using congested roads should be based on the congestion costs expected after
the prices have been introduced. The problem is to judge what effect different
prices will produce on the volume of traffic and hence on the cost of road use.

7.2 Thus it is necessary to estimate the cost of road use for different traffic
speeds on each road or area of roads and to judge the reaction of traffic to
different prices. The estimates of cost used here (see Appendix 2) are made
from data of (a) the speeds normally achieved for different volumes of traffic
on the particular road or area under consideration; (b) the costs borne by
each type of vehicle at different speeds including the value of their occupants’
time; and (c) the composition of traffic, i.e. the proportion of different types
of vehicle. Traffic studies are needed to extend some of this information and
also to determine the existing traffic speeds and flows. Given this information,
there is no difficulty in obtaining realistic cost estimates. Accurate ways of
estimating the reaction of traffic to different prices have, however, not yet
been developed; they may be developed by a process of carefully controlled
trials.

7.3 With this information, price assessments can be made for every road or
group of roads in a congested area. The normal pattern will be for the highest
prices to occur in the centre of a town, declining as one moves away from the
centre. If a continuous pricing system is adopted, price assessment should
be made for areas large enough to be considered as zones. If only one zone
is contemplated in a town, the larger the zone the lower will the price normally
be. If two zones are contemplated, an inner and an outer zone, the prices
in both zones will fall as the inner zone is made larger. If a point pricing
system is adopted, the best prices will be judged according to the conditions
in the surrounding area.

7.4 The ability to adjust the effect of prices by moving the zone boundaries
is important because under the meter schemes put forward only a few prices
would be available to meet the needs of every town and city in the country.
As a result of price calculations made for London and Cambridge the following
scale of prices is suggested * together with possible colour indications : —

PURPLE 2d. per minute
GREEN Id. „

PINK 0.5d. „

BROWN 0.2d. „

BLUE O.ld.

The blue rate is intended mainly as a parking rate (6d. per hour). The brown
rate could serve both as a low running rate and as a high parking rate. The
availability of two parking rates would be a considerable advantage in helping
to divert some of the parking demand from the more popular to the less

* The calculations indicate that ‘ optimum ’ prices might sometimes be higher than 2d. per
minute. There is much to be said, however, for pitching prices on the low side.

Printed image digitised by the University of Southampton Library Digitisation Unit

popular parking sites and thus maintain a better balance between demand and
supply both in an area as a whole and in different parts of the area.

7.5 From time to time price changes, up or down, might be required. In all
the meter systems described above such changes need cause no difficulty pro-
vided that the ratios of the five prices to each other remain the same.

7.6 We are aware that a great deal of theoretical argument may be advanced
on the questions of what prices to charge and how to calculate them. But as
in other sections of the economy a good deal about prices can be learned by
trial and error. Prices can be adjusted until they give the best results. We
would warn, however, against the use of prices to reduce traffic to an arbitrary
level. The ‘ best ’ level of traffic varies greatly from one urban area to another.
From an economic point of view the only satisfactory way to know what the
best level of traffic is in any particular area is to work out the costs in the way
described above and try to set a price just high enough to bring the volume of
traffic down to the level at which the costs imposed by vehicles are equal to the
price charged. Our calculations suggest that considerable variations from
the ‘ optimum ’ price may be made with little loss of benefit.

8. ECONOMIC CONSEQUENCES OF DIRECT CHARGING
The Overall Benefits

8.1.1 The basic effect of applying any of the direct charging systems that
we have described would be to reduce traffic in congested areas and hence
to increase the average speeds and lower the costs (before tax) of the remaining
vehicles. If the prices were fixed according to the method described above
(and demonstrated in Appendix 3) average speeds might rise to 14 mile/h
in areas where they are now only 10 mile/h, and to 17 mile/h where they are
now only 14-15 mile/h.C^) According to data collected in central London,
the reductions in traffic flow necessary to produce these liigher speeds are
20-25 per cent where the present speed is 10 mUe/h and 15-20 per cent
where it is 14-15 mile/h. There would be little purpose in extending direct
pricing to roads where traffic speeds are already over 20 mile/h.

8.1.2 The net benefits to the community resulting from higher traffic speeds
can be classed as follows : —

(a) Savings in the paid working time of persons who travel in working
hours, including the crews of commercial vehicles and buses.

(b) Other time savings, including time travelling to and from work,
shopping, etc.

(d) Greater productivity from buses and commercial vehicles insofar as
the same number of journeys can be made by fewer vehicles, requiring
less capital investment.

(e) Losses to people who as a result of the price changes refrain from
making journeys which they would otherwise have made.

Printed image digitised by the University of Southampton Library Digitisation Unit

(f) Gains and losses to other road users such as pedestrians and cyclists.

(g) Changes in the costs caused by accidents. It is not known how a
reduction in congestion would affect either the number of accidents
or their severity.

(h) Changes in the costs imposed by road users on the rest of the com-
munity by way of fumes, noise, dirt and other undesirable features of
motor traffic.

8.1.3 The five classes (a) to (e) are to some extent measurable and,
according to the estimates (Appendix 3), would yield net benefits of between
£100 million and £150 million a year if a meter system of the sort described
were adopted under present traffic conditions. The total, of which about
7 per cent would accrue in Central London*, is comprised approximately

as follows : —

(a) Savings in paid working time: -f 40 per cent

(b) Other time savings : +63 „ „

(d) Capital savings : +10 „ „

(e) Losses: — ^20 „ „

The remaining three classes, if they were measured, would probably add
significantly to the total benefit. With each year that passes, if congestion
continues to grow wider and deeper, the potential benefits from improved
pricing will rise at a steeper rate than the rise in vehicle-mileage.

8.1.4 The magnitude of the benefits would depend to some extent on the
type of meter system adopted and how it was used. Up to a point, the more
refined systems may yield greater benefit than the simpler systems, but of
course they are bound to cost more. We would not expect the cost of
running the system to be more than 5-10 per cent of the measurable benefits.
The estimates assume a fairly simple system; with a rrlore refined system the
benefits might possibly be higher than the upper estimate.

8.1.5 There are places, such as certain ‘ bottleneck ’ areas, where the
presence of congestion would not in itself justify the introduction of a pricing
zone, or where prices should be deliberately kept low in the interests of,
say, local traffic. There are places where suitable pricing zones might be
difficult to define because of the diffuseness of congestion. There is also
no doubt that under any system the prices charged would often be inappro-
priate for the prevailing traffic conditions and that road users would not
always respond predictably to them. Such inefficiencies are inevitable and
we hope that the estimate of benefits allows sufficiently for them.

8.1.6 The benefits of a new pricing system would be fully realised only
after a period of adaptation. Those who stood to gain would need time
to adjust their habits to the new conditions. But more important, those who
stood to lose would need time before they found — or were offered — the best

* Central London is taken to mean the area covered by the R.R.L. London Traffic Surveys.

Printed image digitised by the University of Southampton Library Digitisation Unit

alternatives and adjusted their habits so as to minimise their losses. Some
people would travel at cheaper (i.e. less congested) times, some would travel
by cheaper (i.e. less congested) routes. Some commercial operators, such
as owners of dehvery vans, might make fewer journeys but with larger
payloads. Shoppers might make fewer but bigger shopping expeditions.
Commuters would be induced to share their cars more than at present.
Some people would transfer to buses and railways. Some might walk or
take a bus for the last part of their journey. Some, but probably a small
minority, would decide that their journey was not worthwhile.

The Effect on Incomes

8.2.1 Changes in the method of pricing the roads would inevitably cause
changes in the distribution of real income, that is to say, some people would
benefit more than others and some people would lose. The way in which
different groups of people were affected would depend largely upon the
manner in which the revenue from the pricing system was used. For
instance, unless steps were taken to avoid it, there might be a change of
relative income between town and country. There might be a transfer of
income from road-users to non-road-users. Special concessions might be
necessary for some road users, and additional restraints for others, e.g.
expense account motorists might need special attention.

8.2.2 There are, of course, innumerable ways in which the revenue could
be used. It could be used simply as general revenue, in which case there
would be a transfer of income from affected vehicle users to the rest of the
community. It could be distributed between local authorities in a manner
calculated to compensate residents in areas where road prices were highest.
It could be used to subsidise public transport, thus compensating those
forced from private to public transport. It could be used to reduce the
existing motoring taxes, in particular those such as the annual licence fee
which do little to restrain the use of congested areas. It is clearly outside
our terms of reference to suggest how the revenue should be used. We
can only point out, therefore, that the consequences of road pricing, like those
of any other charge or tax, cannot be fully assessed until the answer to this
question is known.

Private Motorists

8.3.1 At present motorists are discouraged from entering congested areas
by three forms of restraint; (1) price, as represented by fuel tax and parking
charges, (2) congestion, and (3) the difficulty of finding a suitable parking
place for a sufficiently long period. Under a road pricing system the first
of these three restraints would be substantially increased and the other two
considerably reduced. Motorists in these areas would thus find driving
conditions better but more expensive, and some motorists — while con-
tinuing to use these areas — would on balance consider themselves worse off
than before. It should be observed, however, that, if the volume of traffic
could be controlled thus by price, there would be no reason why adequate
off-street parking should not eventually be provided in congested areas.
The savings in time and convenience which adequate parking facilities can
bring to the motorist have not been included in our estimate of benefits.

Printed image digitised by the University of Southampton Library Digitisation Unit

8.3.2 The main way, however, in which motorists as a class could gain
would be if it were decided that the revenue from road prices paid by them
to the Exchequer should be used to reduce the existing taxes on motorists.
For instance, if the money were returned in the form of reduced annual
licence fees or purchase tax, they would benefit from direct lump sum com-
pensation. Or, if fuel tax were reduced, they would benefit every time they
motored outside charging zones, say on rural journeys, and outside charging
hours, such as during the evening or at weekends.

JPublic Transport

8.4.1 One effect of raising road prices in towns would be to increase the
numbers travelling by rail commuter services. Greater crowding of these
services would impose losses on existing rail passengers. We cannot analyse
here the full consequences for railways, but clearly there might be a need
to review fares, schedules and investment programmes in the light of possible
changes in road prices.

8.4.2 The effect of road pricing upon bus services would be, by reducing
congestion, to improve both their speed and their regularity. As with the
railways, some increase in the number of passengers could be expected.
It is not clear how much buses should be required to pay in the way of road
charges. With the assistance of the London Transport Board we have
examined the effect of road pricing upon bus finances, and hence upon
fares, under a variety of different assumptions. (”) Reduced congestion
can bring considerable savings in the operating costs of bus companies,
largely because the same service can be provided by fewer buses and crews.
If buses were to pay road prices, the correct charging factor would need
careful study, as also would the redistributive effects of the system upon bus
finances and upon peak and off-peak travellers. There is no necessary
reason why road pricing should cause increases in fares. Under some
policies it could lead to lower fares.

Revenue

8.5.1 It is difficult to predict how traffic would react to the price system which
is being discussed, and hence it is difficult to predict with any accuracy
what general level of prices would finally materialise. We would judge, but
we may be far from correct, that the revenue from the price system in urban
areas would amount to £300- £400 million per year under present traffic
conditions. But. whatever this figure happened to be, the revenue could
easily be adjusted within large margins by changes in existing motor taxes
or in the urban road prices themselves, without seriously detracting from
the benefits of the system.

8.5.2 Looking ahead, as vehicle-ownership increases, the pressure of demand
should tend to force road prices up, but the progress of road improvements
and traffic management should increase the supply of road space and hence
tend to bring prices down. In the earlier years, when the growth of vehicle-
ownership is likely to be fastest, we would expect prices and revenue to rise.
Later on prices and even revenue may fall. It has to be realised that invest-
ment in new and better roads, while producing an increase in community
benefit, must often produce a fall in revenue.

Printed image digitised by the University of Southampton Library Digitisation Unit

Effect on City Development

8.6.1 It is sometimes suggested that the imposition of higher charges for
the use of congested streets would result in cities losing their attractiveness as
centres of population, industry and commerce. The question is : would the
restraint of traffic by high charges have a more adverse effect on cities than
would restraint by congestion or by other means? We can only indicate
some of the effects which an increased charge for road use might have on
city traffic. These effects will vary considerably according to the kind of
charging system used.

8.6.2 The occupancy of cars would tend to rise and there would also be a
shift from private cars to buses. Occupations associated with large road
requirements would tend to leave city centres and to make way for occupations
requiring less road usage. Road-users who place a high value on their
time would tend to replace those who put a low value on theirs. For
example, the proportion of mileage run for business and professional purposes
would probably rise. Under an improved pricing system traffic would
become less ‘ peakish i.e. more uniformly distributed both over time and
over the city area. For example, deliveries of goods in central areas would
occur to a greater extent outside the hours of peak congestion.

8.6.3 Some of these effects would help to offset the loss of vehicle-mileage
caused by higher road prices, and the number of people entering the city
might not decline despite a fall in the number of vehicles. The increased
efficiency in the use of city streets should increase, rather than decrease, its
attractiveness. We cannot follow through the implications of these effects,
but we can see no a priori reasons for believing that they would, in general,
be harmful to city development.

Investment Criteria

8.7 The present system of road taxes gives little information about the demand
for road space at different levels and compositions of flow. It is very difficult
to estimate what people would be ready to pay for road improvement, if they
are not, in fact, asked to pay. One of the advantages of road pricing is that
it would give some guidance as to the places where the need for road improve-
ment is greatest.

9. RECOMMENDATIONS FOR FURTHER WORK

9.1 If it is decided to continue the technical development of meter schemes
there is much research in allied fields of traffic, economics and sociology which
could usefully be carried out at the same time. There are six subjects which
should, if possible, be explored further before any detailed pricing proposals
are made. They are :

(1) The congestion costs caused by different kinds of vehicles in different
traffic conditions; including comparisons between different sizes of cars.

(2) The theory of prices and price zones, with case studies.

(3) The comparative advantages of charging by time, distance or points.

Printed image digitised by the University of Southampton Library Digitisation Unit

(4) The possible income redistributive effects of urban road pricing.

(5) Plans for operational and introductory trials.

(6) The effect of road pricing on the cost of commercial transport.

9.2 Other subjects which merit further study and investigation are:

(7) The evaluation of costs, especially those relating to pedestrians and
cyclists and the valuation of unpaid time.

(8) The proportion of through-traffic in towns of different kinds.

(9) The proportion of through journeys spent in congested areas.

(10) Estimates of demand for road space by different user-classes at different
price levels, based on present data and including an enquiry into the
conditions under which travellers change from one form of transport
to another.

(1 1) Estimates of benefits obtainable from different systems of direct pricing.

(12) The effect of road pricing on the operations of buses, taxis and hire cars.

(13) The effect of road pricing on railway congestion.

<14) The effect of road pricing on retail trade and the development of shopping
centres.

10. SUMMARY AND CONCLUSIONS

Section

reference

10.1 We have tried to answer the question: can a method of
charging for the use of the roads be found to serve effectively the
role of the price system? We have pointed out the deficiencies in

this respect of the present methods of taxation, notably their inability 4

to restrain people fiom making journeys which impose high costs
on other people, and we have suggested that road charges could
usefully take more account than they do of the large differences
that exist in congestion costs between one journey and another. 2

10.2 We have examined a number of possible charging methods,

including new methods of charging directly for movement on the
roads. We have found little advantage in two measures sometimes
proposed: the differential fuel tax, which could not be related at 5.1

all closely to congestion costs, and the poll tax on employees in 5.4

congested areas, which — ^whatever its merit in other fields — would

have little effect on road congestion.

10.3 We believe that the parking tax could bring significant 5.2
benefits, in spite of its inequitable results and its undesirable effects

in encouraging non-parking traffic and penahsing local traffic.

Preferable to a parking tax might be a system of daily licences, 5.3
which would embrace all traffic in the areas concerned but would
in turn give rise to difficult boundary problems.

Printed image digitised by the University of Southampton Library Digitisation Unit

Section

reference

10.4 Whatever the merits of the parking tax and the daily licence,
considerably superior results are potentially obtainable from direct
pricing systems. By charging more when costs are high and less
when costs are low, it is estimated that a practicable system in
urban areas could yield economic benefits from reduced congestion

of £100 million to £1 50 million a year under present traffic conditions ; Appx. 3
and this estimate excludes some important items which cannot be
measured.

10.5 We have examined a number of proposals for direct charging
methods, and we have described six meter systems — two manual 6
and four automatic — which, with development, could probably be
made effective. When we started our work, we set out a list of 17
requirements which we considered desirable for a road pricing
system. Some at least of the six meter systems show promise , of
satisfying all these requirements. The main conclusion that
emerges from onr work, therefore, is that there is every possibility

that at least one of these proposals could he developed into an
efficient charging system and could yield substantial benefits on
congested roads.

10.6 We have made no attempt to decide which of the six systems
is likely to be the best. A choice between them cannot be made
until further development has been carried out by the manufacturers.

The necessary development would probably take about a year or
eighteen months from the time that the Government gives the
necessary encouragement. If the technical development is to be
continued there are a number of other allied subjects which we have 9
recommended for investigation at the same time.

ACKNO WLEDGEMENTS

We wish to acknowledge technical assistance received from the Electrical
Research Association, the National Chemical Laboratory, the National Physical
Laboratory, Pye Telecommunications Ltd., General Precision Systems Ltd.,
Joseph Lucas Ltd., Venner Ltd., Ferranti Ltd., Decca Radar Ltd., Ever Ready
Battery Co., General Electric Co., George Kent Ltd., Mallory Battery Co.,
Nash & Thompson Ltd., and the Stewart Aeronautical Supply Co. We are
also grateful for advice and information from the London Transport Board
and the Metropolitan Police; from Mr. W. B. Reddaway, Director of the
Department of AppUed Economics, Cambridge University, and Professor William
Vickrey of Columbia University; and from various members of the Ministry of
Transport and the Road Research Laboratory. Finally we would thank
Mr. J. M. Thomson for acting as secretary to the Panel.

Printed image digitised by the University of Southampton Library Digitisation Unit

REFERENCES

(1) MUNBY, D. L. Road and Rail Track Costs. Manchester Statistical Society
November, 1962.

(2) ROTH, G. J. Economic Benefits to be Obtained from Road Improvements

with Special Reference to Vehicle Operating Costs. Department of Scientific
and Industrial Research, Road Research Laboratory Research Note No. RN/3426
1959, Harmondsworth (Unpublished). ~ ’

(3) WALTERS, A. A. Track Costs and Motor Taxation. Journal of Industrial
Economics, April 1954.

(4) VICKREY, W. Statement to the Joint Committee on Washington Metro-
politan Problems, Nov. 1959.

(5) TANNER, J. C. Pricing the Use of the Roads; a Mathematical and Numerical
Study. Department of Scientific and Industrial Research, Road Research
Laboratory Note No. LN/319, 1963, Harmondsworth (to be published).

(6) THOMSON, J. M. Road Pricing in Central London. Paper PRP.I8, 1962,
Harmondsworth (Road Research Laboratory) (Unpublished).

(7) THOMSON, J. M. Calculations of Economic Advantages Arising from a
System of Road Pricing. Paper PRP.8, 1962, Harmondsworth (Road Research
Laboratory) (Unpublished).

(8) CHARLESWORTH, G. and J. L. PAISLEY. The Economic Assessment of
Returns from Road Works. Proc. Instn. Civ. Engrs., 1959: 14 (Nov.), 229-54.

(9) FAIRTHORNE, D. Possible Effects of Staggering Hours of Work in Centi'al
London. Department of Scientific and Industrial Research, Road Research
Laboratory Note No. LN/205, 1962, Harmondsworth (Unputjlished).

(10) EATON, J. E. and F. V. WEBSTER. Traffic Observations in London During
and After the Bus Strike in 1958. Department of Scientific and Industrial
Research, Road Research Laboratory Research Note No. RN/3306, 1958,
Harmondsworth (Unpublished).

(11) WARDROP, J. G. Some Theoretical Aspects of Road Traffic Research.
Proc. Instn. Civ. Engrs., Part II, 1952, 1 (2), 325-362.

(12) FOSTER, C. D. Two Notes on Some of Mr. J. M. Thomson’s Calculations.
Paper PRP.34, 1963, Harmondsworth (Road Research Laboratory) (Un-
published).

(13) ROTH, G. J. Pricing the Use of Roads: Potential Gains Available in
Cambridge. Department of Applied Economics, University of Cambridge,
Nov. 1962 (Unpublished).

(14) TANNER, J. C., H. D. JOHNSON and J. R. SCOTT. Sample Survey of the
Roads and Traffic of Great Britain. Department of Scientific and Industrial
Research, Road Research Technical Paper No. 62. H.M.S.O. (1962).

Printed image digitised by the University of Southampton Library Digitisation Unit

(15) EATON, J. E. and JUNE I. BRYANT. London Traffic Survey No. 9,
Department of Scientific and Industrial Research Road Research Laboratory.
Research Note No. RN/3965, Apl. 1961, Harmondsworth (Unpublished).

(16) BUCHANAN, JAMES M. The Pricing of Highway Services. National Tax
Journal, June 1952.

(17) THOMSON, J. M. Economic Effects of Road Pricing upon the London
Central Bus Fleet. Paper PRP.26, 1962, Harmondsworth (Road Research
Laboratory) (UnpubUshed).

(18) FAIRTHORNE, D. The Location of Charging Points in Central London.
Paper PRP.33, 1963, Harmondsworth (Road Research Laboratory) (Un-
published).

(19) BECKMANN, M., C. B. McGUIRE and C. B. WINSTEN. Studies in the
Economics of Transportation. Yale University Press, 1956, Ch. 4.

(20) WALTERS, A. A. Optimum Motor Taxes. Applied Statistics, 1961.

(21) BEESLEY, M. E. and G. J. ROTH. Restraint of Traffic in Congested Areas.
Town Planning Review, Oct. 1962.

(22) DAY, A. C. L. Roads. Mayflower Press, 1963.

(23) SMEED, R. J. The Traffic Problem in Towns. Manchester Statistical Societv.
1961.

(24) WALKER, G. Prospect for Transport in the 1960’s. Westminster Bank
Review, May 1960.

(25) WALTERS, A. A. The Theory and Measurement of Private and Social Cost
of Highway Congestion. Econometrica, Vol. 29, No. 4 (1961).

(26) REYNOLDS, D. J. Planning, Transport and Economic Forces. Journal of
the Town Planning Institute, Vol. XLVII, No. 9 (1961).

(27) ROTH, G. J. A Pricing Policy for Road Space in Town Centres. Journal of
the Town Planning Institute, Vol. XLVII, No. 9 (1961).

(28) ROTH, G. J. and J. M. THOMSON. Road Pricing: a cure for Congestion?
Aspect, Vol. 1, No. 3, 1963.

(29) REYNOLDS, D. J. and R. F. F. DAWSON. The Economics of Toll Roads.
Department of Scientific and Industrial Research, Road Research Laboratory
Research Note No. RN/37S8, 1960, Harmondsworth (Unpublished).

(30) BLACKBURN, A. J. The Problem of Urban Traffic Congestion. Hertford
College, University of Oxford, 1963 (Unpublished).

(31) FOSTER, C. D. The Transport Problem. Blackie, 1963.

Printed image digitised by the University of Southampton Library Digitisation Unit

APPENDIX 1*

Note on Marginal Cost Theory and Road Pricing
1 . Marginal Cost

1 . 1 Consider a particular road which permits vehicles to undertake a specific journey
from a point X to another point Y. The service which the road supplies (including
such lifting, traffic control, cleaning, maintenance, etc. as are provided) is the
facility for vehicles, together with their passengers and other contents, to make the
journey from X to Y.

1.2 We make the following simplifying assumptions:

(a) all vehicles make the same journey;

(b) all vehicles using the road are identical and are driven in an identical manner
so that in any given set of conditions the costs associated with their use are
identical;

(c) the influence upon traffic speed of all factors other than the number of vehicles,
such as pedestrians or weather conditions, or the spacing of vehicles, is constant
and may be regarded as one of the permanent characteristics of the road
determining its speed/flow relationship.

1.3 If there is no tax of any sort, and if n vehicles per hour make the journey in
t minutes each, the costs incurred are

n f (t) + r (n) + s (n)

where f (t) represents the operating costs of each vehicle, as a function of time t,
and r (n) represents the road maintenance costs and s (n) represents the other social
costs (such as the effects of noise, dirt and fumes). It may be noted that these costs
do not include the original cost of building the road or any other costs which are
not affected by the volume of traffic which now uses the road.

1.4 If the flow of vehicles increases to n+1 per hour and the journey time for each
vehicle consequently rises from t to t^ the total costs increase to (n+1) f (t^) -i-
r (n+l) + s (n+I). Thus the increase in total cost is

(n+1) f (t 1) - n f (t) + r (n+l) ~ r (n) + s (n+l) — s (n)

This is the marginal cost at (n+l) vehicles per hour.

2. Private and Public Cost

Now for die sake of simplicity, let us assume that the differences r (n+I) — r (n)
and s (n+l) — s (n) are both negligible when n is large, although this assumption
may not always be valid. Then marginal cost equals (n+l) f (t^) — n f (t). This
is the cost that could be avoided if any one vehicle refrained from making the journey.
But the cost actually borne by each vehicle, if there is no tax, is only f (t i) which is
less than the costs it causes by an amount n [f (t i) — f (t)]. This is sometimes called
the difference between private and public cost. According to the theory, unless some
added restraint is exercised journeys will be made which are valued at more than
private cost but less than public cost. A tax may provide the added restraint.

* Prepared for the Panel by the Road Research Laboratory. See also References 19-31.

Printed image digitised by the University of Southampton Library Digitisation Unit

3. Diagrammatic Demonstration

3.1 The theory may be demonstrated by means of a diagram, Fig. 1, as in Beesley
and Roth

per

vehicle

Fig. 1.

Three variables are represented. First, the average cost curve shows the sum of
avoidable vehicle operating and road costs, for a journey, at different volumes of
traffic, i.e. vehicle flows. As the flow increases, after a certain point speed declines
and costs rise. (At some extreme point there is a maximum possible flow and it is
believed that, if the number of vehicles trying to use the road exceeds this maximum,
it can lead to an actual flow less than the maximum. Hence the cost curve turns
back at this point. But this is irrelevant to the argument.)

3.2 Road users pay their own operating costs, which for the sake of simplicity we
have assumed to be equal for all vehicles, and we shall now also assume that they pay
their part of road costs, say by means of a fuel tax. Then, if no additional charge
is made, the price of a journey which each vehicle has to pay is the sum of its own
operating costs and its contribution to road costs, i.e. the amount represented by
the average cost curve. As the price rises, the number of vehicles that will make the
journey declines, and this is represented by the demand curve. The position of
equilibrium is indicated by the point of intersection B, between the demand curve
and the average cost curve, where flow is Q.

Printed image digitised by the University of Southampton Library Digitisation Unit

3.3 The third curve, the marginal cost curve, shows the marginal cost at different
levels of flow, i.e. the increase in total costs caused by the addition of each extra
vehicle. It is evident that, as the flow increases, each extra vehicle adds more to the
total cost than it pays and, when the flow exceeds P, there are some vehicles which
are not willing to pay the costs they cause. If a charge CA reduces the flow of
traffic from Q to P, where the demand curve intersects the marginal cost curve, journeys
will not be made unless they are valued at more than the cost they cause.

4. Variations between Vehicles

4.1 In the above argument we assumed that all vehicles were identical and that
their costs in any given set of conditions were also identical. If we now withdraw
this assumption it is clear that the cost per mile caused by one vehicle can vary con-
siderably from that of other vehicles. If the same price per mile is charged to all
vehicles, some will pay more than the costs they cause and others less. (It is some-
times useful to measure costs and prices in units of time rather than of distance
since a price per minute will usually give a closer approximation to cost than will
a price per mile.)

4.2 The inefficiency of charging all vehicles the same price clearly varies with the
extent to which individual vehicle costs vary. One way of reducing this inefficiency
is to classify the vehicle population (into, say, cars, motorcycles, taxis, lorries of
various sizes, etc.) and to charge each vehicle according to its class. There may be
other reasons too for wanting to charge one class of vehicle more than another.

5. Variability of Other Factors

Our assumptions excluded the effect on journey costs of fluctuations in external factors,
such as pedestrians and weather, and also of fluctuations in the timing of journeys
themselves. The occurrence of these factors can at any time cause journey costs to
be higher or lower than normal. Many fluctuations are unpredictable and un-
avoidable. But many others are predictable and can, in theory, be allowed for in the
setting of prices. For instance, hourly, daily and seasonal fluctuations are largely
predictable both as regards the volume and composition of traffic demand and the
degree of interruption to be expected from pedestrians and cyclists. Nevertheless,
insofar as such fluctuations occur and are not allowed for, further inefficiencies may
occur.

6. Recovery of Total Costs

For reasons which do not concern us it may be necessary to recover total costs,
including original building costs, or to contribute towards general taxation. If
prices are made equal to marginal cost, the total revenue obtained may fall short
of the required amount. It may be necessary to avoid either a surplus or a deficit
and this probably cannot be done without causing a divergence between price and
marginal cost and hence some shift in the use of resources. There are, however,
several different methods of increasing revenue, and some methods cause a smaller
shift than others in the use of resources. A deficit may sometimes be recovped
simply by raising prices, or by price discrimination, or by various kinds of indirect
prices or taxes, the choice of method depending upon the circumstances and, in
particular, upon the complex of demand elasticities. Since in practice there is no
difficulty in recovering the total costs of the road system, we shall not discuss this
aspect further.

7. Conclusion

7.1 We have been careful to distinguish between road costs and vehicle operating
costs, although both groups of costs are equally essential to the production of
journeys. The reason for the distinction is simply that the first group, road costs,

Printed image digitised by the University of Southampton Library Digitisation Unit

are usually paid by the agencies which manage the roads while the second group,
the vehicle operating costs, are usually paid direct by the consumer. This unusual
division of responsibility for the costs causes much confusion about road pricing.
It is helpful to think of a situation similar to that on the railways in which the
‘ management ’ of the road also owns all the vehicles and pays all the drivers. They,
the management, will then need to take account of the marginal cost of journeys'
i.e. the effect of each additional journey on the costs of the whole enterprise, and will
try to prevent journeys taking place which do not yield a return at least equal to
their marginal cost, unless there are special reasons for permitting them.

7.2 The most efficient price system might appear to be one in which price varied
with cost on every road at every moment of the day. But this presupposes that
road users are able and willing to take account of such a highly complicated system.
In practice, of course, they are not. If the price system is complicated road users
will probably find simple ‘ rule of thumb ’ methods to tell them approximately what
the average prices are and roughly what the prices of particular journeys are likely
to be, and they will act accordingly. If this is so the complicated system may be no
more efficient than a simpler system.

7.3 . In conclusion,^ therefore, the argument for making road users pay the costs which
their journeys inflict on others, including other road users, is subject to certain
reservations of which we mention three. One is the requirement regarding total
revenue. Another is the limit to the amount of complexity which in practice can
have any useM effect on the decisions of road users. A third reservation, which is
the main topic of the Panel’s report, is the technical possibility of charging the prices
and collecting the payment.

APPENDIX 1*

Calculation of Marginal Cost

1. Consider a road network on which v represents the average speed of traffic at a
given time, expressed in miles per hour, and q represents the average flow (i.e. the
number of vehicles per hour passing a point) at that time.

Let v = f (q)

The operating cost per mile for an average vehicle is given by c. Variable road cost
IS given by r. All costs are in pence per mile,
c = g (v)
r = h (q)

Total variable cost is given by C, marginal cost by m.

C = cq -T r

dC dc dr

m. = — = c “r cj — -f- —

dq dq dq

We know that — is relatively small and in the absence of data we assume it to be a
dq

constant R.

Let marginal public cost, i.e. the cost inflicted by one road user upon other road
users, be denoted by m'.

m' = m — c — R
dc

= q —
dq

• Prepared for the Panel by the Road Research Laboratory.

Printed image digitised by the University of Southampton Library Digitisation Unit

It has been demonstrated by Charlesworth and Paisley (8) that c may be expressed
in the form

c = a + - (a, b constants)

It has been shown by Thomson! *) (following Wardrop(“)) that for a network, as
for a single road, v may be expressed in the form

V = d — • f q (d, f constants)

dc b (d-v)

It follows that m' = q — =

It remains to put values to b and d. Dawson’s estimate for 1962, b = 230 pence,
includes leisure time at three-quarters the value of working time; composition of
traffic is taken as 68 per cent cars, 4 per cent buses, 14 per cent light commercial,
14 per cent other commercial. For the constant d, we take Thomson’s 1961
estimate for Central London, d = 28.

230 (28 — V)

Thus m' = —

In Table 1 values of ra' are given for speeds between 5 and 20 mile/h. There are
dangers in extrapolating too far and for this reason the values of v < 8 raile/h should
be viewed with caution. To translate from marginal public cost to the appropriate
price at any given speed, bearing in mind that the speed must be that obtaining after
the imposition of the price, one must add something for R, which may be about
one penny per mile.

Table 1

Marginal Public Cost

Traffic speed
(mile/h)

Marginal public cost
(pence per mile)

212

140

164

13-3

10-8

8-8

7-1

5-7

4-6

2. The estimates in Table 1 apply to an ‘ average ’ vehicle. The costs caused by
a car would normally be a little less. These estimates are based on data which
applies to London off-peak traffic. Obviously the parameters vary from one town
to another and from one time to another. The most important difference between
different places might be the average width of streets. Estimates show, however,
that the width of street makes little difference to the marginal cost at any given
speed.

Printed image digitised by the University of Southampton Library Digitisation Unit

APPENDIX 3 *

Estimate of Economic Benefits from Direct Pricing

1 Several sets of calculations have been made, for a nation-wide system ( ^)( ^ -), for

Central London(^), and for Cambridge(i^). Here we give an example of the basic
method adopted. Minor variations of method exist and the calculations can be carried
to different levels of refinement.

2 Consider a road network in which the flow is q and the price per mile is p, p being
understood to denote all costs borne by the vehicle drivers and passengers, including
the cost of their time. The flow is related to the price by a demand curve of the form
q=ffp). Alternatively this may be expressed as the inverse relationship p = f'Hq)
where f*^ f(p) = p.

Before the introduction of any price change, let the price be a and the flow be Q.
It may be assumed that when the price is p, the “^oss” benefit to every userj of the
network (including the price of making the journey) is at least equal to p. By integrating
over all road users, it may be deduced that the gross benefit to all road users is

f-i (q) dq

From this must be subtracted the total price aQ giving for the net benefit (or consumer's
surplus) to all road users

f-i (q) dq— aQ {])

Now suppose that an additional charge per vehicle-mile of 0 is imposed. When the
charge is added the flow declines, giving rise to higher traffic speeds and a consequent
reduction in costs from a to a' (let a-a'=g). Let the new flow be Q'. The gross benefit
to all road users becomes

rQ'

f-i (q) dq and the total price is (a' 4-0) Q'. Therefore the net benefit or con-
o

Sumer’S surplus is
rQ^

f-i (q) dq— (a' + 0) Q'.
o

However, the amount 0Q' is a transfer payment and represents no real cost to the
community. Therefore the real net benefit is

fQ'

f-i (q)dq— a'Q' (2)

Subtracting (1) from (2) gives the increase (or decrease if negative) in net benefits
resulting from imposition of the charge 0.

The increase G in net benefits is

G = aQ— a' Q'— f‘‘ (q) dq (3j

Jq'

* Prepared for the Panel by the Road Research Laboratory.

X Assuming that there are no road users apart from vehicle users.

Printed image digitised by the University of Southampton Library Digitisation Unit

If it is assumed that when the flow is of the order of Q or Q' the demand function is
k k k

q =f(p) so that the elasticity of demand is unity, then Q=~, and

a a'+3

p=f'^ (q)=-. By substituting in equation (3) it can readily be deduced that the

increase in net benefits is

G=PQ' + aQ log —

and therefore
G Q' Q'

— =P — h alog —

Q Q Q

where — is the ratio of the new flow to the old flow.

(4)

(5)

3 In order to quantify the above relationship a process of inspection is necessary
Q' Q'

because — varies with fi. First the relationship between g and — must be established
Q Q

This follows from the speed-cost relationship and the speed-flow relationship.

The speed-cost relationship has been taken here as

13'92

(6)

where a is the average cost in shillings for one passenger car unit travelling at v mile/h.
This is the formula from Charlesworth and Paisley(S) brought up to present-day
values and converted to a p.c.u. basis. The speed-flow relationship has been taken as

v=28—

125

(7)

where q is the flow in p.c.u’s per hour and v is the speed in mile/h. This is Thomson’s
formulae for Central London.

4 From these two formulae we obtain Table 2, which gives the value of g for various

flow reductions — from five initial situations in which the base-speeds selected

are 8, 10, 12, 14 and 16 mile/h.

Printed image digitised by the University of Southampton Library Digitisation Unit

Table 2

Table of g and Q'jQ

•o _ O'

100-0

91-7

83-3

75-0

66-7

0-058

0-112

0-158

0-201

T3 ^ a

y ^ C5

« V o

■u C

100-0

92-8

85-7

78-6

71-4 1

64-3 '

57-1

pq'^

0-077
0-144
0-202 !
0-256
0-302 i
0-345

*3 ^ O'

w ~ o

&= "
fli s

100-0

93-7

87-5

81-2

75-0

68-7

62-5

56-2

50-0

Tt-fSO\'O^OOrJ-f^

66606000

-a ^ O'

IS i

i'S

100-0
94-4
88-9
83-3
77-8
72-2
66-7
61-1
1 55-6

50-0
44.4

«o

0-147

0-268 ;

0-372 ;

0-460

0-537

0-604

0-662

0-716

0-762

0-805

m O'

B=5. g

O.JJ 2

^ 2

100-0 i

95-0
90-0
85-0
80-0
75-0
70-0
65-0
60-0
55-0
' 50-0

45-0
40-0

5 OO
PQ

0-246

0-424

0-578

0-692

0-796

0-884

0- 961

1- 029
1-086
M40
1-186
1-229

Average private ’
costperp.c.u.-mile
(shillings)

2-328

2-082

1-904

1-757

1-636

1-532

1-444

1-367

1-300

1-242

1-188

M42

1-099

Journey

speed

mile/h

.52

Printed image digitised by the University of Southampton Library Digitisation Unit

5 Values of Q7Q can now be found for different values of a and p. The net benefits
G

per vehicle-mile, — , resulting from a charge p shillings per vehicle-mile, then follow

from formula (5). Some results are shown in Table 3.

Table 3

Net Benefits from Different Charges at Different Speeds
(in pence per p.c,u.-mile)

Initial

traffic

speed

mile/h

3d.

6d.

9d.

1/-

l/3d.

l/6d.

l/9d.

2/-

— Benefits —

2-0

3-6

4*7

5-5

6-0

5-9

5-5

1-4

2-4

3-4

3-6

3-7

3-4

2-9

2-1

1-2

2-0

2-1

2-2

2-1

1-4

0-7

-0-4

1-0

1-5

1-7

1-3

1-0

0-5

—0-4

—1-8

. 0-8

0-9

0-8

0-3

—0-4

—1-1

—2-0

—2-8

The benefits in bold indicate the optimum prices at each of the five speeds.

6 It has been estimated that the number of motor vehicle-miles run in urban adminis-
trative areas in 1959/60 was nearly 37,000 million.(^'^) There is very little information
of the speeds at which this mileage was travelled. It is assunied that one-half of this
mileage occurs in areas which become charging areas. Various hypothetical speed
distributions can then be considered. For instance, if the mileage all occurred at 16
mile/h and a rate of 6d. per mile were charged, the benefits would be about £70 million.
This may be regarded as a minimum estimate.

7 Estimates by Tanner, Thomson and Foster, using a variety of assumptions, have
yielded figures of between £109 million and £213 million. A fair estimate probably
lies within the range £100 to £150 million.

APPENDIX 4*

Benefits from Parking Taxes

Fixed Parking Tax without Non-parking Traffic or Peak Flows
1. (a) Reduction of traffic

Consider a road network in which a parking tax of p per p.c.u. is irnposed on cve^
journey, non-parking traffic being ignored. Since the tax per p.c.u. journey is fixed,
the tax per mile depends on the length of journey. If the journey length is t miles,
the tax is p/t per p.c.u.-mile.

* Prepared for the Panel by the Road Research Laboratory.

Printed image digitised by the University of Southampton Library Digitisation Unit

Now suppose that the demand function for journeys of length between t and t +
dt is

f(t)dt
qt dt =

where pt is the price per p.c.u. mile (including the cost of time and the tax if any),
e is the elasticity of demand and f(t) is a function of t. Let the cost per mile,
excluding tax, be reduced from a to a' as a result of the reduction in traffic when
the tax is introduced. Then if Qt dt and Q't dt are the p.c.u. mileages travelled on
journeys of length between t and t + dt before and after the introduction of the
parking tax, respectively,

f(t) dt f(t) dt

Qt dt = — and Q't dt = (1)

(a' + 3)“
t

Integrating over all journey lengths gives the. total p.c.u. mileages Q and Q' before
and after introduction of the tax.

/-oof(t) dt

- and Q'

00 f(t) dt

(a' +J)'
“ t

The simplifying assumption is now made that f(t) is constant say, K, for journey
lengths ti to ta, and zero otherwise, that is to say, every journey length interval
between 1 1 and tz possesses the same demand function (in terms of vehicle-mileage)
and journeys whose lengths are outside this range are ignored.

(t2— ti)K r

Then Q = and Q' = K

(a' H-_P)'
t

a= r dt

tz-ti J (a'-l-J)'

ti t

.( 2 )

(3)

2. (b) Net benefits

The consumer surplus before imposition of the fax is
tz Qt

p. dq — a'Qt

ti o

and when the parking tax is imposed this is altered to

pt dq — (a' -I- 3 ) Q't ^ dt

Printed image digitised by the University of Southampton Library Digitisation Unit

However, the term — Q't is a transfer payment which does not affect the net change
t

in total benefits. The net gain or net increase in benefits is therefore

t2 Qt

G =

(a Qi — a' Q't) dt —

pt dq dt

ti Q't

= oQ — a' Q'-

t2 Qt

ti Q't

Pt dq dt

If q = — , if 1 1 < t < t2 and q = 0 otherwise, then
Pt=

G = a Q — a' Q' —

r ;

K Te

Q't

q i

)

\t2-t 1

dq dt .

..(4)

3. (c) Formulae for particular elasticities

By substituting various values of the elasticity of demand e in equations (1), (2),
(3) and (4), formulae can be derived for the relative vehicle-mileages and benefits.
The following results have been found.

If e = 1

Q a'

G a

Q t2 — 1 1

Ife =

Q' Va

Q a'(t2 — ti)

a' (t2 — ti)

- log

a' t2 -b 3
a' 1 1 -b 3

a' ti + 3 a' t 2 + 3

ti log t 2 log

Utl Ut2

Vt 2 (<x' t 2 + 3) — V 1 1 (o' 1 1 + 3)

3 Va' t2 -b 3 + V«' t2

log ^ 7—

Vet Va' ti + 3 -b Va' ti

Printed image digitised by the University of Southampton Library Digitisation Unit

G 2Va

— = 2a H

Q t2— ti

Vti (a' ti + P) — v't2(«' t2 + P)

}

If e = 2

Q'_ ^
Q a'2

G a

Q ' a'

2P a' t 2 + 3

1 log

3 "

a' (t 2 -t i) a' + 3 (a' 1 1 + P) (a' t 2 + 3) J

a 32

{a' 1 1 + 3 ) (a' t2 + 3 )

Tfte Introduction of Non-parking Traffic into the Model
(a) Reduction of traffic

4. LetQ = Qp + Qn : (1)

and Qn = or Q

where Qp, Qn are the volumes of parking and non-parking traffic expressed in
p.c.u.-miles. This describes the starting position with no parking tax and price
equal to a. After the imposition of a parking tax p equation (1) becomes

= Q'p -r Q'n

Let the demand function for non-parking traffic be
K'

Qn = — (K' constant)
pe

K' = a»Qn

a= Qn a‘ B Q

After imposition of tax, Q'n = =

a' ® a' '

Qp'

and Q'p = (1 — o) Q .

Q' Qp' a' o

= (1 - cr) 4-

Q Qp a' '

We know the value of Q'p/Qp in terms of a, a', ti, ta and 3, for the different
elasticities of demand, e= 4, e = 1, e = 2.

(b) Net benefits

5. The total gain G is the sum of the gains Gp and Gn to parking and non-parking
traffic respectively.

Printed image digitised by the University of Southampton Library Digitisation Unit

Thus

G Gp Gn Gp Gn

= 1 = (1 — a) h a

Q Q Q Qp Qp

By similar reasoning to that given in 14.2,

Gn = uQn — Ct ^ Q^n +

TQ'n

p dq

J Qn

Gn a

If e=l, — = a log —

Qn a'

G Gp a

Hence if ev^l, — = (1 — a) h ci<t log —

Q Qp 0.'

Gn U

If e=l, — =

Qn 1 — 0

Hence if e

(1— a) h 2ac (1—

G Gp aa(a — a')

and if e = 2, — = (1 — n) 1

Q Qp a'

have already been calculated in terms of a, a', ti, tz and

Numerical Evaluation

(a) a fixed parking tax without non-parking traffic or peak flows

6. The most simple, basic situation is that in which there is no non-parking traffic,
and the volume of parking traffic is constant throughout the day, or throughout the
period of control. In the formulae obtained for calculating the net benefit that would
result from a fixed parking tax, the only variable about which nothing is known is the
elasticity of demand. The elasticity of demand is the proportionate effect that a change
in price will have upon the volume of demand. Thus when the elasticity is -t, a price
increase of x per cent produces a fall in demand of iix per cent ; when the elasticity is 2,
the same price increase produces a fall in demand of 2x per cent. Elasticities of -J-,
1 and 2 are considered and thus a fairly wide range of possibilities is allowed for.

7. The first results are shown in Table 4 and refer to vehicle mileage evenly distributed
on journey lengths of between 1 and 20 miles, with no non-parking traffic and no peak
flow. Six possible situations are considered, in which the elasticity of demand is low,
medium and high (j, 1 and 2) and in which the base-speed (i.e. the average journey
speed before tax) is alternately 10-7 mile/h, which is the current speed in Central
London,(i 5) and 8-0 mile/h, which is the speed to which traffic might descend if no
permanent cure for congestion is found. The table shows the effect of parking taxes
ranging from 3s. to 15s. upon the volume of traffic, and the net benefits in pence per
vehicle-mile (i.e. per vehicle-mile travelled before tax).

Printed image digitised by the University of Southampton Library Digitisation Unit

Table 4

Elfect of a fixed parking tax:
(without non-parking traffic or peak flows)

Base-speed

Parking

tax

(shillings)

Perce

elastic!

itage rec
n traffic
ty of den

uction

land is:

Net benefits
(pence per
veh/mile)

(b)I

(a)

(b)

(0)

10-7 mile/h

6*5

9-3

11-6

1-5

1-9

2-2

8-3

12'3

15-5

1*7

2-1

2-4

10-1

15-0

19-3

1-8

2-2

2-5

11-9

17-6

23-3

1-9

2-2

2-6

13-5

20-2

27-0

1-9

2-2

2-4

15-1

22-7

31-0

1*9

2-1

2-2

16-6

25-1

34-2

1-8

1-9

2-0

17-9

27-4

37-6

1-7

24-3

37-7

47-0

0-8

0-2

8-0 mile/h

4-1

5*7

6-7

2-2

2-7

2-9

5-6

7-5

8-8

2-7

3-2

3-4

7-0

9-3

ll'O

3-1

3-6

3-8

S-3

13-3

3-4

3-9

4-2

9-5

12-9

15-6

3-6

4-2

10-6

14*7

17'9

3-7

4-3

4-6

11-7

16-5

20-3

3-8

4-4

4-7

12-8

18-3

22-7

3-8

4-4

4-7

13-9

20-1

25-2

3-9

4-4

4-7

15-0

21-8

27-7

3-9

4-3

4-6

161

23-5

30-2

3-8

4-2

4-5

17-1

25-2

32-6

3-7

4-0

4-2

18-1

26-9

35*0

3-6

3-9

4-0

8, Some interesting points arise from the results in Table 4. The exact level of the tax
makes little difference to the benefits. A difference of a few shillings either side of the
optimum makes a big difference to the level of congestion but very little difference to
the net benefits. In other words after a certain initial amount of traffic has been forced
off the roads there is a wide range over which the losses roughly equal the gains
Eventually, of course, as the tax is raised higher and higher the losses exceed the gains!

9. Because of the way in which the benefits “flatten out” at the top, the elasticity of

demand becomes of little consequence. The same tax gives the optimum results for
each of the three elasticities studied. When the base-speed is 10-7 mile/h the best tax
IS clearly 6s; when the base-speed is 8-0 mile/h the best tax is 11s. Moreover the elasti-
city of demand makes little difference to the benefits. The base-speed however has a
big influence upon the benefits. ’

10. In the optimum positions for the six situations under consideration the net
benefits vary from l-9d. per veh/mile to 4-7d. per veh/mile.

(b) A variable parking tax

^ ''“able par^g tax means that motorists pay according to the number of hours
that they are parked. It is assumed in our model that there is no correlation between
the nurnber of hours parked and the length of journey within the area. A variable tax
IS then the same as charging a fixed tax, T, to one-hour parkers and a fixed tax 2T
to two-hour parkers, and so on. The maximum benefits will be obtained when the
average tax is at or near the optimum for a fixed tax but they will obviously be lower
than under a fixed tax since the tax will be far higher tlian the optimum for long
parkers and far lower for short parkers.

Printed image digitised by the University of Southampton Library Digitisation Unit

12. This argument against the variable tax applies in all circumstances unless parking
time is highly correlated with length of journey, which is unlikely to occur in practice,

13. We now introduce non-parking trafBc into the model. The appropriate formulae
have been derived above and an example of the results is shown in Table 5. In this
example two of the six situations used in Table 4 are used again to show the effect of
non-parking trafBc. The proportion of non-parking trafBc is shown before and after
the tax and each situation is considered with three different levels of non-parking
traffic. For each situation the optimum tax has been calculated, i.e. the tax which
produces maximum benefits, and the benefits are compared with what they were
before the introduction of non-parking traffic.

Table 5

Effect of non-parking traffic

Base-

speed

mile/h

Elasticity

demand

Percentage

non-parking

traffic

Optimum

tax

Percentage
reduction
in traffic

Net benefits
with and without
non-parking
traffic
pence per
veh/mile

Before

tax

After

tax

With

Without

10-7

13-5

13-2

1-9

2-2

33-0

12-1

1-6

2-2

604

8-1

1-3

2-2

8-0

15-7

15-3

3-9

36-3

124

62-3

7-0

1-9

14. The conclusions to be drawn from Table 5 are that the presence of non-parking
traffic (a) has little effect on the optimum parking tax; (b) leads to a higher degree of
congestion at the optimum; and (c) causes a considerable reduction in the benefits
to be obtained from the policy. In the two examples shown the elasticity of demand = 1 .
If it were lower than this the effect of non-parking traffic would be less marked, and
vice versa. There is no reason to suppose, of course, that the elasticity of demand for
non-parldng traffic will be the same as for parking traffic. The worst situation will be
where the former is high and the latter is low so that the tax leads to a small reduction
in parking traffic which is almost entirely offset by the increase in through-traffic.

(d) Estimate for London

15. We shall now make our model resemble London as closely as possible. Vehicle-
mileage has already been taken to be evenly distributed on journeys of length between
1 and 20 miles, and this is thought to be a reasonable approximation for London
conditions. The relationship between flow and operating costs is known quite well.
The proportion of non-parking traffic has been assumed to be 30 per cent. The peak
flow in Central London is known to be only about 10 per cent above off-peak, although
it is much more pronounced in suburban areas. The biggest gap in our knowledge is the
elasticity of demand. This is assumed to be 1 in the off-peak and J in the peak.

16. The incidence of congestion is not uniform throughout London and therefore a
policy of parking restriction would not be applied throughout the city, but only in the
centre and in some suburban centres. It is difficult to say whether the unevenness
of congestion tends to make our estimates too high or too low. The estimates are given

Printed image digitised by the University of Southampton Library Digitisation Unit

in Table 6 and show the effect of a tax upon the volume of trafSc dming both peak and
off-peak tours, and also the value of the net benefits. Different values of tax are con-
sidered between 3s. and 15s.

n The figures in Table 6 indicate that the best parking tax would be 6s. This tax
would reduce traffic by about II per cent in the off-peak period and by 7 per cent m
ffie peak, and speeds would rise from 10-7 mile/h o 12-6 mile/h, and from 9-8 mile h
to 1 H tole/h respectively. The real cost of travel m London would go down by 12
per int but the real savings (after taking the losses into account) would be about n
per cent or 1-5 pence per original vehicle-mile.

Table 6

A parking tax in London

Tax

Off-peak hours

Peak hours

Net
benefits
(pence per
veh/mile)

Speed

Reduction
in traffic
per cent

Spe

Reduction
in traffic
per cent

Before

After

Before

After

10-7

11-8

6-5

9-8

10-5

3*9

1-2

10-7

12-1

8-1

9-8

10-7

1-3

10*7

12-4

9-6

9-8

10-9

6-2

1-4

10-7

12-6

9-S

11-1

7-2

1-5

10-7

12-9

12-5

9*8

11-3

8-1

1*5

10-7

13-1

13-7

9-8

11-4

9-0

1-4

10*7

13-3

14-9

9-8

li-6

9-8

1-3

10-7

13-5

16-0

9-8

11*7

10-5

1-2

10-7

14-3

20-9

g's

2 2-3

13-9

0-5

18 The effect of a parking tax in London would therefore appear to be of modest
value. To put the matter in perspective one may compare it with an efficient road
pricing solution. Both, the effect on speed und the net benefits of the parking policy
are about 40 per cent of what is claimed for a pricing solution.( S)

APPENDIX 5*

Differential Fuel Taxes: fuel-fetching Journeys

5.1 If it is assumed that drivers fill only their tanks, fuel-fetching journeys will not
be worth making if the tax differential per gallon per mile is less than

2 X Total Vehicle Operating Cost per Mile

Tank capacity

2 X fuel price per mile (excl. tax)

Tank capacity

2 X fuel price per gallon (excl. tax)

Tank capacity x fuel consumption
(m.p.g.)

2 X fuel price per gallon (excl. tax)

Range of vehicle in miles

* Prepared for the Panel by the Road Research Laboratory.

Printed image digitised by the University of Southampton Library Digitisation Unit

Total operating cost per mile
Fuel price per mile (excl. tax)

Ratio of total operating cost per mile
to fuel price per mile (excl. tax)

5.2 Thus, whether it would pay drivers to make fuel-fetching journeys would depend
on the tax differential, on the range of their vehicles, and on their operating costs,
including an evaluation of their own time and inconvenience.

5.3 Excluding petrol tax, the average operating cost of cars has been estimated at
just over three times the basic cost of petrol. Drivers’ evaluation of time and incon-
venience varies both between individuals and for any given individual from one time
to another. One might expect that drivers would tend to make fuel-fetching journeys
in time on which they placed a low value, provided that this coincided with the times
when their cars needed refuelling. It is, therefore, thought that most drivers would
probably estimate their total operating costs (including the cost of time and incon-
venience) at between, say, 5 times and 8 times their basic tax-free petrol costs (assuming
journey speeds of less than 30 mile/h).

5.4 The average price of petrol, excluding tax, is about ls.9d. per gallon.

5.5 Most British cars have a range of between 180 and 300 miles, but since it is not
usually possible to judge exactly when tanks will run dry, it would be reasonable to
assume ranges of between 150 and 300 miles. If a oar has a range of 300 miles and the
driver estimates his total operating costs at 5 times the tax-free petrol cost, then fuel-
fetching journeys to tax-free areas will be profitable if the tax differential exceeds

2 X ls.9d. X 5

per gallon per mile

300

or 0-7 pence per gallon per mile.

5 6 If a car has a range of 150 miles and the driver estimates his total operating costs
at 8 times the tax-free petrol cost, then fuel-fetching journeys to tax-free areas will
be worth while if the tax differential exceeds

2 x ls.9d. X 8

per gallon per mile

150

or 2-24 pence per gallon per mile.

5 7 If the fuel to be fetched is subject to some tax, the above figures should be
increased slightly, e.g. if the tax is Is. per gallon, the dffierential of 0-7d. s^hould be
raised to 0-8d.; if the tax is 2s. per gallon, the differential should be raised to 0-9d.

5 8 If we assume that the tax would never fall below Is. per gallon, we conclude that,
if the differential did not exceed Id. per gallon per mile, few drivers would make fuel-
fetching journeys, whereas if the differential was more than 2d. per gallon per mile,
fuel-fetching journeys would probably become profitable for most drivers. ^

5 9 However even if the tax differential was below Id. per gallon per mile, some
drivers could be expected to make fuel-fetching detours from their normal routes, if
these involved less additional mileage and inconvenience than journeys for the sole
purpose of fuel-fetching.

5 10 A further factor to be considered is that the speed of a fuel-fetctog journey
has been assumed as less than 30 mile/h. With the advent of motorways fuel-fetchmg
could be made much faster and hence more profitable for many motorists, indicating
a lower limit to the differential than that suggested above.

Printed in England for Her Majesty’s Stationery Office by Liverpool Printing & Stationery Co„ Ltd.
Wt. 5084/K.52/PS 307770. 4/64. (82081) ^ = s

Printed image digitised by the University of Southampton Library Digitisatiori Unit

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Road pricing : the economic and technical possibilities"

See other formats